Module1: Introduction

The power industry across the globe is experiencing a radical change in its business as well
as in an operational model where, the vertically integrated utilities are being unbundled and
opened up for competition with private players. This enables an end to the era of monopoly.
Right from its inception, running the power system was supposed to be a task of esoteric
quality. The electric power was then looked upon as a service. Control consisting of planning
and operational tasks was administered by a single entity or utility. The vertical integration
of all tasks gave rise to the term vertically integrated utility. The arrangement of the
earlier setup of the power sector was characterized by operation of a single utility
generating, transmitting and distributing electrical energy in its area of operation. Thus,
these utilities enjoyed monopoly in their area of operation. They were often termed as
monopoly utilities.

Why were earlier utilities the monopolies'? The reason for monopoly can be traced right
back to the early days when electricity was comparatively a new technology. The skeptical
attitude of the government towards electricity led to investment by private players into the
power sector, who in turn, demanded for the monopoly in their area of operation. This
created a win-win situation for both- government and the electrical technology promoters.
However, the government would not let the private players enjoy the monopoly and exploit
the end consumer and hence introduced regulation in the business. Thus, the power
industries of initial era became regulated monopoly utilities . The structure of a conventional
vertically integrated utility is shown in Figure 1.1. As evident from the figure, there was only
a single utility with whom the customer dealt with. Thus, only two entities existed in the
power business: a monopolist utility and the customer.

Fig 1.1

What does regulation mean? The regulations are generally imposed by the government or
the government authority. These essentially represent a set of rules or framework that the
government has imposed so as to run the system smoothly and with discipline, without
undue advantage to any particular entity at the cost of end consumer. All practical power
systems of earlier days used to be regulated by the government. This was obviously so. The
old era power industries were vertically integrated utilities and enjoyed monopoly in their
area of operation. Whenever a monopoly is sensed in any sector, it is natural for the
government to step in and set up a framework of way of doing business, in order to protect
end consumer interests. Some of the characteristics of monopoly utility are:

1. Single utility in one area of operation enjoying monopoly.

2. Regulated Framework: The utility should work under the business framework setup
by the government.
3. Universal Supply Obligation (USO): Utility should provide power to all those
customers who demand for it.
4. Regulated Costs: The return on the utility's investments is regulated by the
government.

In a nutshell, regulation is about checking the prices of the monopolist in the absence of
private players and market forces.

Reasons for restructuring / deregulation of power

industry
The next obvious question is, what is deregulation or restructuring of an industry? From
the name, one can sense discontinuation of the framework provided by the regulation. In
other words, deregulation is about removing control over the prices with introduction of
market players in the sector. However, this is not correct in a strict sense. An overnight
change in the power business framework with provision of entry to competing suppliers and
subjecting prices to market interaction, would not work successfully. There are certain
conditions that create a conducive environment for the competition to work. These
conditions need to be satisfied while deregulating or restructuring a system. Sometimes, the
word deregulation may sound a misnomer. Deregulation does not mean that the rules
wont exist. The rules will still be there, however, a new framework would be created to
operate the power industry. That is why the word deregulation finds its substitutes like re-
regulation, reforms, restructuring, etc. The commonly used word in Europe is
liberalization of power industry; deregulation is a more popular phrase in US.

If the power industries worked successfully with the regulated monopoly framework for over
100 years, what was the need for deregulating or changing the business framework of the
system? There are many reasons that fuelled the concept of deregulation of the power
industry. One major thought that prevailed during the early nineties raised questions about
the performance of monopoly utilities. The takers of this thought advocated that monopoly
status of the electric utilities did not provide any incentive for its efficient operation. In
privately owned utilities, the costs incurred by the utility were directly imposed upon the
consumers. In government linked public utilities, factors other than the economics, for
example, treatment of all public utilities at par, overstaffing, etc. resulted in a sluggish
performance of these utilities. The economists started promoting introduction of a
competitive market for electrical energy as a means of benefit for the overall powerector.
This argument was supported by the successful reform experiences of other sectors such as
airlines, gas, telephone, etc.

Another impetus for deregulation of power industry was provided by the change in power
generation technology. In the earlier days, cost-effective power generation was possible
only with the help of mammoth thermal (coal/nuclear) plants. However, during the mid
eighties, the gas turbines started generating cost effective power with smaller plant size. It
was then possible to build the power plants near the load centers and also, an opportunity
was created for private players to generate power and sell the same to the existing utility.
This technology change, supposed to have provided acceleration to the concept of
independent power producers, supported the concept of deregulation further. This
technology change is supposed to have provided acceleration to the concept of independent
power producers. This further supported concept of deregulation. This was specifically true
where the financial losses were apparently high which was prevalent in some of the
developing countries.

It should be noted that these are the indicative or major reasons for introducing the concept
of deregulation in power industry. There are many other reasons as well. One of the
important reasons is the condition under which power systems were regulated, did not exist
any more. There was no wind of skepticism about the electrical technology and all the initial
investments in infrastructure were already paid back. Further, the deregulation aims at
introducing competition at various levels of power industry. The competition is likely to
bring down the cost of electricity. Then, the activities of the power industry would become
customer centric.

The competitive environment offers a good range of benefits for the customers as well as
the private entities. It is claimed that some of the significant benefits of power industry
deregulation would include:

1. Electricity price will go down: It is a common understanding that the competitive

prices are lesser than the monopolist prices. The producer will try to sell the power
at its marginal cost, in a perfectly competitive environment.
2. Choice for customers: The customer will have choice for its retailer. The retailers will
compete not only on the price offered but also on the other facilities provided to the
customers. These could include better plans, better reliability, better quality, etc.
3. Customer-centric service: The retailers would provide better service than what the
monopolist would do.
4. Innovation: The regulatory process and lack of competition gave electric utilities no
incentive to improve or to take risks on new ideas that might increase the customer
value. Under deregulated environment, the electric utility will always try to innovate
something for the betterment of service and in turn save costs and maximize the
profit.

The deregulation of the industry has provided electrical energy with a new dimension where
it is being considered as a commodity. The commodity status given to electrical power has
attracted entry of private players in the sector. The private players make the whole business
challenging from the system operators point of view, as it now starts dealing with many
players which are not under its direct control. This calls for introduction of fair and
transparent set of rules for running the power business. The market design structure plays
an important role in successful deregulation of power industry.

Understanding the restructuring process

The process of deregulation has taken different formats in different parts of the world. Also,
the reasons for power sector to adopt the reforms vary from country to country. For the
developed countries, introduction of competition to achieve social welfare was probably the
most important reason. On the other hand, the developing countries mainly banked on the
capacity addition through entry of private players. It is observed that neither, there is lone
reson for driving deregulation of power industry nor is there a single objective of the same.

The restructuring process starts with the unbundling of the originally vertically integrated
utility. This essentially leads to separate the activities involved in an integrated power
system leading to creation of functional partition amongst them. For example, the
unbundling of power industry involves separating transmission activity from the generation
activity. Further, distribution can be separated from transmission. Thus, these three
mutually exclusive functions are created and there are separate entities or companies that
control these functions. Then, the competition can be introduced in the generation activity
by allowing other private participants in this segment. In contrast to the vertically
integrated case where all the generation is owned by the same utility, there is a scope for
private players to sell their generation at competitive prices. The generators owned by the
earlier vertically integrated utility will then compete with these private generators. The
transmission sector being a natural monopoly is most unlikely to have competing players in
the sector. This is because for natural monopolies like transmission companies, the business
becomes profitable only when output is large enough. Figure 1.2 shows the representative
structure of deregulated power system. In contrast to the vertically integrated utility
structure, it can be seen that there are many alternative paths along which the money
flows. It is evident that there are many more other entities present, apart from the
vertically integrated utility and the customers. It should be noted that there can be many
more versions of deregulated structure.

Various Entities Involved in Deregulation:

The introduction of deregulation has introduced several new entities in the electricity market
place and has simultaneously redefined the scope of activities of many of the existing
players. Variations exist across market structures over how each entity is particularly
defined and over what role it plays in the system. However, on a broad level, the following
entities can be identified:

1. Genco (Generating Company): Genco is an owner-operator of one or more

generators that runs them and bids the power into the competitive marketplace.
Genco sells energy at its sites in the same manner that a coal mining company
might sell coal in bulk at its mine.
2. Transco (Transmission Company): Transco moves power in bulk quantities from
 where it is produced to where it is consumed. The Transco owns and maintains the
transmission facilities, and may perform many of the management and engineering
functions required to ensure the smooth running of the system. In some
deregulated industries, the Transco owns and maintains the transmission lines
under the monopoly, but does not operate them. That is done by Independent
System Operator (ISO). The Transco is paid for the use of its lines.
3. Discom (Distribution Company): It is the owner-operator of the local power
delivery system, which delivers power to individual businesses and homeowners. In
some places, the local distribution function is combined with retail function, i.e. to
buy wholesale electricity either through the spot market or through direct contracts
with Gencos and supply electricity to the end use customers. In many other cases,
however, the Discom does not sell the power. It only owns and operates the local
distribution system, and obtains its revenue by wheeling electric power through its
network.
4. Resco (Retail Energy Service Company): It is the retailer of electric power.
Many of these will be the retail departments of the former vertically integrated
utilities. A Resco buys power from Gencos and sells it directly to the consumers.
Resco does not own any electricity network physical assets.
5. Market Operator: Market operator provides a platform for the buyers and sellers
to sell and buy the electricity. It runs a computer program that matches bids and
offers of sellers and buyers. The market settlement process is the responsibility of
the market operator. The market operator typically runs a day-ahead market. The
near-real-time market, if any, is administered by the system operator.
6. System Operator (SO): The SO is an entity entrusted with the responsibility of
ensuring the reliability and security of the entire system. It is an independent
authority and does not participate in the electricity market trades. It usually does
not own generating resources, except for some reserve capacity in certain cases. In
order to maintain the system security and reliability, the SO procures various
services such as supply of emergency reserves, or reactive power from other
entities in the system. In some countries, SO also owns the transmission network.
The SO in these systems is generally called as Transmission System Operator
(TSO). In the case of a SO being completely neutral of every other activity except
coordinate, control and monitor the system, it is generally called as Independent
System Operator (ISO).
7. Customers: A customer is an entity, consuming electricity. In a completely
deregulated market where retail sector is also open for competition, the end
customer has several options for buying electricity. It may choose to buy electricity
from the spot market by bidding for purchase, or may buy directly from a Genco or
even from the local retailing service company. On the other hand, in the markets
where competition exists only at the wholesale level, only the large customers have
privilege of choosing their supplier.

Understanding the restructuring process

Electricity, as a commodity, can not be compared with any other commodity traded in the
market. This is because it has some distinguishing characteristics of its own, which demand
satisfaction of technical constraints before accomplishing the commercial trades. Two
important features of electricity as a commodity are: need for real time balance and inability
to wheel the commodity through desired path (in bulk). Hence, a set of principles laid down
by standard micro-economic theory can not be mapped directly to the electricity commodity
markets.
Tackling network congestion is one of the challenging issues of the de-regulated era.
Transmission network provides the path through which transactions are made in a power
market. But each transmission network has its own physical and operating limits like line
flow limits, bus voltage magnitude limits and more. The power injection and withdrawal
configuration should be such that no limit gets violated. If the network is operated beyond
these limits, it may, even, result in the entire system blackout. Therefore, any arbitrary set
of transactions cant be organized on the power network. This has given rise to a new
problem under the restructured power system environment, referred to as congestion
management. There are many ways in which congestion is formally defined but to explain in
simple words, when some components in a power network appear to be overloaded due to a
trading arrangement, that particular arrangement is said to create congestion on the
network. The purpose of congestion management is to make necessary corrections in order
to relieve congestion. It can be easily appreciated that under the vertically integrated
structure, network congestion, in fact, is not a challenging task. This is because all the
resources in the system are under the direct control of the monopolist. Thus, this is the sole
responsibility of the monopolist to maintain its transmission network.

Provision of ancillary services is another tough task carried out by the system operator
under the deregulated framework. Ancillary services are defined as all those activities on
the interconnected grid that are necessary to support the transmission of power while
maintaining reliable operation and ensuring the required degree of quality and safety. Under
the deregulated power system environment, the system operator acquires a central
coordination role and carries out the important responsibility of providing for system
reliability and security. It manages system operations like scheduling and operating the
transmission related services. The SO also has to ensure a required degree of quality and
safety and provide corrective measures under contingent conditions. In this respect, certain
services, such as scheduling and dispatch, frequency regulation, voltage control, generation
reserves, etc. are required by the power system, apart from basic energy and power
delivery services. Such services are commonly referred to as ancillary services. In
deregulated power systems, transmission networks are available for third party access to
allow power wheeling. In such an environment, the ancillary services are no longer treated
as an integral part of the electric supply. They are unbundled and priced separately and
system operators may have to purchase ancillary services from ancillary service providers.

Then, there are certain issues like market design and market power which need regulatory
intervention. Issues pertaining to market design revolve around choice made in the
selection of dispatch philosophies, choice of various pricing schemes, choice between
number of markets with multiple gate closures, etc., from various alternatives. The market
architecture, which maps various markets on timeline, is also an important sub-topic of
market design process.

Existence of market power shows the signs of deviation from the prefect competition. In
general, market power is referred to as ability of market participants to profitably maintain
the market price above or below the competitive level for a significant period of time. To
tackle the situation, an indirect regulatory intervention in the form of market design rules is
needed. Thus, as mentioned earlier, deregulation does not mean ceasing to have rules. It is
the restructuring of the power business framework. More rigorous treatment to these
issues is given in further chapters.

Reasons and objectives of deregulation of various

power systems across the world
Restructuring or deregulation is a broad term and can have different meanings in different
countries. This is because the changes essential for betterment of power sector depend on
the prevailing conditions in the power sector of respective countries. Further, the word
betterment can be looked upon subjectively. For example, well developed, industrialized
countries can expect price to go down and these countries can treat the change in the prices
as betterment. On the other hand, the developing countries need to make radical changes
in the policy and regulation such that barrier to entry for private players is removed. The
effective betterment can be looked upon from this perspective for developing countries.

In this section we will see, in brief, the issues that led to restructuring of the power industry
for following regions / countries: US , UK , Nordic Pool and developing countries.

The US

The US electric utilities, from the very beginning were privately owned and worked in a
vertically integrated fashion. The developed countries like US had well functioning and
efficient electricity systems. However for some systems, so long as consumers were
concerned, they were not satisfied with the rising costs of electricity. For some other
systems, utility management found that running the system was not viable due to low tariff.
In some systems, pressure from smaller players to open up the business for competition
played a major role. By and large, deregulation took place in developed countries by
pressure to reduce costs while simultaneously increasing competitiveness in the market.

Existence of market power shows the signs of deviation from the prefect competition. In
general, market power is referred to as ability of market participants to profitably maintain
the market price above or below the competitive level for a significant period of time. To
tackle the situation, an the indirect regulatory intervention in the form of market design
rules is needed. Thus, as mentioned earlier, deregulation does not mean ceasing to have
rules. It is the restructuring of the power business framework. More rigorous treatment to
these issues is given in further chapters.

The UK

The transformation of the British power sector proceeded along three paths in 1990. First,
the traditional industry was unbundled both vertically and horizontally. High-voltage
transmission assets were transferred to a new National Grid Company (NGC). Coal and oil
fired units were divided among two companies National Power and PowerGen. Nuclear
Electric retained control of all nuclear units. At the outset, National Power had 52 percent of
total generating capacity, PowerGen had 33 percent, and Nuclear Power had the remaining
15 percent. The second set of changes involved ownership. Both National Power and
PowerGen became private companies in 1991, whereas the difficulties associated with
nuclear power resulted in continued government ownership of all nuclear units.
Approximately 30 percent of shares in National Power and PowerGen were sold to the
public,an equal amount to foreign and institutional investors. The remaining 40 percent was
held by the government until 1995. The third set of changes sought to open the system to
competition, wherever possible, while continuing necessary regulations. Vertical and
horizontal restructuring of power generation was based on the assumption that generation
had become workably competitive and would become increasingly so with new market
entrants.

A report on reform process was floated by the regulator in 2001 which stated that wholesale
electricity prices had not fallen in line with reductions in generators input costs and that a
lack of supply side pressure and demand side participation; and inflexible governance
arrangements had prevented reform of the arrangements.

The Nordic Pool

The reforms in Nordic countries were inspired by the electricity market reforms in England
and Wales in 1989, as well as by widely held beliefs that increased competition would raise
power industry efficiency to the benefit of consumers. Norway was first amongst the Nordic
countries to liberalize its electricity market in 1991, but without privatization. The
Norwegian electricity sector remains almost entirely in public hands. Rather than implement
national reforms, the other Nordic countries chose to reform by merging with the existing
Norwegian market, Sweden joining the expanded Nordic pool in 1996, Finland in 1998 and
Denmark in 1999.

The Developing Countries

The case of developing countries is different from that of other countries. In these countries,
the electricity supply is treated as a social service rather than a market commodity. The
ownership of the power sector in these countries is directly under the governments of
respective countries. These state owned-controlled systems have led to the promotion of
inefficient practices over a period. The power sectors of these countries are marked by
supply shortages. There has been an inability to add to the generating capacity. The
subsidies and high transmission and distribution losses are the major concerns before these
systems. Another consequence of state control over electric utilities was the high level of
overstaffing.

The inability to raise funds for capacity addition invited financial support from international
financial institutions like World Bank. These institutions mandated opening of the power
sector for private companies which were contracted under build, own, operate and transfer
(BOOT) scheme.
 Module2:Fundamentals of Economics
The commodity market started with the barter system and eventually developed to the era of
electronic trading systems. In simple words, a market is defined as a meeting place for buyers and
sellers to strike a deal. The technological advances have progressed to such an extent that now-a-
days, a deal for a commodity can be struck with a click of a mouse. This is the concept of a virtual
market place that has the potential of eliminating a physical market place. Microeconomics is the
branch of economics that deals with how households or firms make decisions and how they interact in
the markets. The restructured power systems treat electric energy as a commodity rather than a
service as in vertically integrated systems. As in the case of any other commodity, the behavior of
consumer and suppliers in electricity markets are analyzed using the concepts of microeconomics. This
chapter is aimed at providing some fundamental concepts associated with microeconomics that are
relevant to electricity market. We start by modeling consumer behavior, followed by supplier behavior
modeling. Later on we will see how these behaviors set equilibrium at the marketplace.

Consumer Behaviour

Total Utility and Marginal Utility

It is a common practice to explain concepts of microeconomics using common consumer goods as an

example. However, we will explain it by considering number of units of electrical energy as a
commodity. This would help the reader to relate things later on while studying concepts associated
with electric power markets.

Total utility and marginal utility

We begin with the notion that the consumer achieves some satisfaction from consuming a product,
electric energy in this case. If this satisfaction is absent, the consumer would not demand it at all. This
term is called total utility. Similarly, marginal utility is the utility obtained from the last unit consumed.
Let us explain these terms with the help of an example.

Let there be a square shaped room as shown in Figure 2.1. Small circles in the figure represent
incandescent lamps located at the same height from the ground, as a source of light for the whole
room. Suppose there are nine lamps in all, fitted in the room as per the plan shown in the figure. Let
us assume that all lamps are of the same rating and illuminance (a scientific word for brightness).
When all lamps are off, there is complete darkness in the room. Let us assume that there is an
interlock arrangement in the switchgear such that for putting lamps ON at L level, the lamps at M level
should be ON. Similarly, for putting lamps at M level ON, the central lamp - C should be ON. Now
suppose a person enters the room and puts lamp C ON. Since this lamp is at the center, it spreads
even light all across the room. This light is good enough for a person to move to each and every
corner of the room. Let us assume that the satisfaction' this person gets by putting lamp C ON is 10
units.

Now suppose the person switches the lamp - M1 ON. This bulb still adds to the brightness of the room,
but the satisfaction that it adds to the person is lesser than that provided by lamp C. For this lamp,
the person in the room may not get a satisfaction of 10 units, but it will definitely be lesser than 10
units as room is already lit from the state of total darkness by lamp C. Let us say the person gets
satisfaction equal to 9 units. Similarly, for all lamps sequentially put ON thereafter would render
diminishing satisfaction to the person in the room. This satisfaction for each of the lamps lighted is
nothing but the marginal utility. For first lamps C, it was 10 units, for subsequent lamps it went on
reducing. Let us define total utility as the sum of marginal utilities.
 Figure 2.1:Room with lamp positions

The results of marginal and total utility gained by putting lamps on sequentially are tabulated in Table
1

No. of Marginal Total

Lamp
Lamps utility utility
- 0 - 0

C 1 10 10

M1 2 9 19

M2 3 8 27

M3 4 7 34

M4 5 6 40

L1 6 5 45

L2 7 4 49

L3 8 3 52

L4 9 2 54

Table 2.1: Marginal and total utility after putting the lamps on
Law of Diminishing Marginal Utility

The above pattern of marginal utility provided by sequence of putting bulbs on is called as law of
diminishing marginal utility. It states that after consuming a certain amount of a good or service, the
marginal utility from it diminishes as more and more is consumed. This law is quite natural and should
hold for most of the products one consumes.

Consumer Surplus

The person entering the room in the previous example would have been indifferent to the number
lamps to be put ON had electricity been for free. Then the person would not have bothered about the
marginal utility and the total utility. But as soon as the person is made to pay for the usage of electric
energy, he would start thinking and would rather make a judicious choice about how many lamps to
put on. The person then would have calculated how much utility he could have obtained if he had
spent same amount of energy on other usage, for example say, air conditioner. In other words, how
many lamps the person would have put ON depends not only on the marginal and total utilities but
also on the price of electricity.

Let the price of electricity be INR 4 for each lamp equivalent rating. We now define marginal utility of
one INR as the extra utility when additional one INR is spent on other available usage in general. For
the sake of simplicity, let it be 1 unit. In other words, after spending one INR, the marginal utility
associated with it is one unit. Having the information on price and marginal utility of INR, the person
can determine how many lamps to be put ON. Consider only one lamp, C is ON. The person obtains
marginal utility of 10 (from table 2.1). Since, marginal utility of 1 INR is equal to 1 unit of utility, the
utility obtained in this case would be (10/1 =10 INR). On the expenditure side, the person spends 4
INR to get the lamp ON. Then, the person will go ahead and put the lamp ON (as a matter of fact, he
is left with no choice in this case!). When the person puts the second lamp ON, i.e., M1, he obtains a
marginal utility of 9 units which is equivalent of (9/1) = 9 INR. Since, the marginal utility that the
person would obtain is greater than the price he pays (INR 4), he would go ahead to put the second
lamp ON. The person keeps on making such comparisons before putting the next lamp ON.

What happens when he puts ON lamp L2, i.e. seventh lamp? It renders utility equal to 4 INR which is
equal to price per unit of electricity. The answer is that the person will be indifferent. However, one
thing is for sure: the person will not put ON lamps after 7th lamp as the price he pays is more than
the worth he accrues. This decision making process can be easily understood with the help of Figure
2.2. The horizontal line at INR 4 depicts the price of electricity. Before lamp no. 7 is lit, the marginal
utility (converted into INR) is more than the price of electricity. At 7th lamp it becomes equal while for
lamps 8, 9 and 10, the marginal utility is lesser than the price.

Table 2.2 summarizes these results at each stage. It provides details about total utility and total
expenditure at each stage. Last column of Table 2.2 depicts nothing but the person's surplus in terms
of money. In other words, it is the difference between what he gains and expenditure towards the
same. The shaded area in Figure 2.2 is called as net consumer surplus. The net consumer surplus
represents the extra value that the person in the room is able to get from being able to buy all the
electric energy at INR 4, even though the value he attaches to electric energy is higher than the price
of electricity.
Figure 2.2:Decision making based on marginal utility and net consumer surplus

Total Total
No. of Surplus
utility Expenditure
Lamps (INR)
(INR) (INR)

0 0 0 0

1 10 4 6

2 19 8 11

3 27 12 15

4 34 16 18

5 40 20 20

6 45 24 21

7 49 28 21

8 52 32 20

9 54 36 18

Table 2.2: Details about total expenditure and total utility at each stage
Consumer Equilibrium

The word equilibrium used in generic terms means the position of balance. In the above example, the
person in the room will stop or rest or attain equilibrium after lighting 6th or 7th lamp on. Last column
of Table 2.2 reveals the fact that the person's surplus is maximized at 6th and 7th lamp. In other
words, the shaded area in Figure 2.2 representing net consumer surplus is maximized at equilibrium.
In general, we can then say that consumer's equilibrium with respect to the purchase of one good is
attained when the difference between total utility in terms of money and the total expenditure on it is
maximized.

Market Demand Curve

The relation shown in Figure 2.2 is termed as individual demand curve. It is easy to infer that the
peculiar nature of this curve is due to law of diminishing marginal utility. Thus, the demand curve can
be termed as marginal utility curve. If, instead of the discrete demand curve as shown in Figure 2.2, a
continuous function is established, the nature of the curve will be the one with negative slope or
downward sloping.

It is unlikely that all the persons entering the dark room will feel the same marginal utility with each of
the lamps. If we aggregate the individual demand curves of sufficiently large number of consumers,
the discontinuities of Figure 2.2 will be smoothened and will give a market demand curve or the
demand function. This is shown in Figure 2.3.

Figure 2.3: Demand function

Demand Elasticity

The downward sloping demand curve can be seen from a different perspective. It emphasizes that a
small increase in the price of a commodity will decrease its demand. The rate of change of the
demand curve with respect to price would surely quantify the change. However, to make the changes
comparable, the percentage changes rather than absolute changes are computed. Thus, the price
elasticity of demand becomes the ratio of relative change in demand to the relative change in price. It
is given as:

......................................................................................................(2.1)
Where depicts elasticity, price and q, the quantity.
The price elasticity of demand can be defined as a measure of how much the quantity demanded of a
good responds to a change in the price of that good, computed as the percentage change in quantity
demanded divided by the percentage change in price.
Downward sloping demand function of Figure 2.3 and equation 2.1 together depict that the price
elasticity of demand will be a negative number. Many a times, elasticity is presented as an absolute
value by dropping a minus sign. Measuring demand responsiveness then becomes a simple task. If the
elasticity number is higher, higher is the demand responsiveness. We follow the convention with
negative sign in-tact. Various cases of price elasticity of demand are established in Table 2.3.

Sr. Elasticity
Type of Elasticity
No. Range

1 =0 Perfectly inelastic

2 -1 < < 0 Inelastic

3 = -1 Unit elastic

4 - < < -1 Elastic

5 =- Perfectly elastic

Table 2.3: Various cases of demand elasticity

Figure 2.4 shows plots of demand elasticity for various cases. In case of unit elasticity, the demand
curve takes shape of a rectangular hyperbola. When demand is perfectly inelastic, the demand curve
takes the form of a vertical line parallel to y axis. It means demand is not responsive to changes in
price.

Supplier Behaviour

Law of Diminishing Marginal Product:

Just as we have law of diminishing marginal utility in case of consumer, we have law of diminishing
marginal product for supplier. This law establishes the input-output relationship for the producer in
short-run. This law is depicted in Figure 2.5
 Figure 2.4: Various cases of demand elasticity

Supply Function

Suppose, the total commodity output is called as y. Let us assume that there is only one factor of
production, x'. Thus, the production function is given as

y = f(x).........................................................................................................................(2.2)

For almost all goods and technologies, the production y increases with x at the beginning. But as
cheaper resources start depleting, costlier resources are employed for production and for the same
quantity of production, the cost starts increasing. In other words, the rate of increase of y decreases
as x gets larger. The inverse of production function will be:

x = g(y).......................................................................................................................(2.3)

This function indicates how much of the variable production factor is required to produce a specified
amount of commodity. If unit cost of factor of production x is w, then, the cost function is given as:

cos t (y)= w. g(y)...............................................................................................................(2.4)

Figure 2.6 shows cost function and the marginal cost function which is the derivative of the cost
function. The convexity of the function is due to law of diminishing marginal product.
 Figure 2.6: Cost function and marginal cost function

Supply functions

Suppose there are many suppliers and they make use of different technologies and fuels to produce
electric energy. Thereby, these producers will have different marginal costs and will have different
power producing quantities at different price levels. If the amount supplied by a large number of
producers is aggregated, a smooth and upward sloping curve is obtained as shown in Figure 2.7. This
is typically known as supply curve.

Figure 2.6: Cost function and marginal cost function

Suppliers' surplus
Suppliers' surplus can be explained on similar lines to the consumer surplus. The entire supply of
commodity is traded at the market price . The suppliers' revenue is the product of traded quantity q
and the market price . Let us consider Figure 2.8. The horizontal line depicts the market price . The
shaded portion shows producers' net surplus. This arises from the concept of all goods being traded at
a price higher than their opportunity costs. Eventually, net surplus is the area between supply curve
and horizontal line depicting market price.

Figure 2.8: Suppliers' net surplus

The supplier whose opportunity cost is equal to the market price is called as marginal producer. The
marginal producer neither accrues any profit nor loss. The infra-marginal producers accrue profit while
the extra-marginal producers would find it worthwhile to sell the commodity only in case market price
increases.

Supplier's Equilibrium
If supplier's marginal cost function is as shown in Figure 2.6, at what level should the supplier stop
producing that commodity? To answer this question, let us explain what is meant by opportunity cost.
The supplier has a threshold price in mind below which it will not sell its commodity. There are two
reasons for deciding this threshold price. First of course is that the total revenue will be less than total
cost of producing that commodity. Second and important reason could be that the supplier could make
use of same resources required to produce the commodity under consideration to produce some other
commodity that would fetch more money. In that case, the revenue from the sale of the commodity
under consideration will be less than the opportunity cost associated with the production of the same.
In other words, the supplier will sell the commodity at a price at which the opportunity cost of
production is equal or lesser.

Supply Elasticity
An increase in the price of a commodity encourages suppliers to make larger quantities of this
commodity available. The price elasticity of supply quantifies this relation. The supply elasticity can be
defined in a similar fashion to the demand elasticity. Only difference is to replace supply curve by
demand curve.
 ......................................................................................................(2.5)

It is worthwhile to note that due to upward sloping nature of supply curve, the price elasticity of
supply will be a positive number. Various cases of price elasticity of supply are shown in Table 2.4.

Sr. Elasticity
Type of Elasticity
No. Range

1 =0 Perfectly inelastic

2 -1 < < J Inelastic

3 = -1 Unit elastic

4 - < < -1 Elastic

5 =- Perfectly elastic

Table 2.4: Various cases of supply elasticity

 Figure 2.9 shows plots of supply elasticity for various cases.

Market Equilibrium

We have seen how consumers and suppliers would behave individually. Let us now consider how the
consumers and suppliers would interact with each other at the marketplace. We take the case of a
perfectly competitive market, although the electricity market is seldom a perfectly competitive market.
A perfectly competitive market has many attributes, the most important being that a single player is
not able to change the market price. Similarly, under perfect competition case, all market buyers buy
at the same market clearing price. The market equilibrium is achieved at a price called market clearing
price such that the quantity that the suppliers are willing to sell is equal to quantity that the
consumers wish to obtain. In other words, market equilibrium is a state of zero excess demand and
zero excess supply.
 Figure 2.10: Market equilibrium

Consider Figure 2.10 for more explanation. It depicts supply and demand curves for a particular
product denoted by S and D, respectively. What should be the market equilibrium price? Suppose that
price is 1. At this price, the consumers demand the quantity d1 and the producers supply the
quantity s1. Obviously, there is a mismatch. Consumers want more than what the producers are
willing to supply. The excess demand will create competition among the buyers and push the price up.
It will increase, say, to 2. Excess demand is present at this price also. Thus price will increase
further. Indeed, the price will keep increasing as long as there is an excess demand. Finally it will
converge to *, at which there is no excess demand. Corresponding quantity is q*.

Just the opposite happens if the initial price is 3 . The quantity demanded d3 is less than the quantity
supplied s3. There is a mismatch in the form of excess supply. The price will keep falling as long as
there is excess supply. Where will the price finally settle? The answer is again * , at which there is no
excess supply.

Thus,* and q* mark the price and quantity at equilibrium, respectively. The equilibrium situation in a
competitive market is said to be Pareto efficient. An economic situation is Pareto efficient if the benefit
derived by any of the parties can be increased only be decreasing benefit enjoyed by one of the other
parties. In Figure 2.10, suppose the quantity exchanged is s1 instead of q*. At that quantity, there is
someone willing to sell extra units of the good considered at price 1, which is less than the price 4
that someone else is willing to pay for that extra unit. If trade can be arranged between these two
parties at any price between 1 and 4 , both parties will be better off and as per definition, this is not
the Pareto efficient situation. Thus, if total amount traded is less than the equilibrium q*, the situation
is not Pareto efficient. Similarly, any amount in excess of the equilibrium value is not Pareto efficient
because the price that someone would be willing to pay for an extra unit is lower than the price that it
would take to get it supplied.

It is not difficult to conclude that the Pareto efficiency is achieved only when goods are allocated on
the basis of a single marginal rate of substitution, as happens in a competitive market.

Areas DS1, DS2, SS1, SS2, etc. in context to the explanation given under "global elfare" and
"deadweight loss" not depcted on figure 2.11

Global Welfare
Global welfare is the sum of net consumer surplus and net producer surplus. It is the quantification of
the overall benefit that arises from trading. Global welfare is maximized when market is settled at the
intersection of supply and demand curves. The global welfare is also termed as social welfare and
social surplus. Global welfare is explained with the help of Figure 2.11. In this figure, sum of the areas
DS1, DS2 and DW2 represents the consumer surplus while sum of areas SS1, SS2 and DW1
represents the producer surplus. The total area consisting of areas DS1, DS2, SS1, SS2, DW1 and
DW2 represents the global welfare. It is clear from the figure that if the price is set to any infra-
marginal value rather than the equilibrium price, there is a reduction in the global welfare.

Figure 2.11: Global welfare and deadweight loss

Deadweight loss

What happens when the price is forcefully set at some value other than the equilibrium price? It leads
to reduction in global welfare and creation of deadweight loss. Suppose the price is set at 2 due to
some intervention by say, the government, as shown in Figure 2.11. In this case, the consumers
reduce their consumption from q* to q. The consumer surplus then becomes equal to area DS1, while
producers' surplus is the sum of areas DS2, SS1 and SS2. Similarly, if price is set at 1 , the suppliers
reduce their production to q from q*. The net consumer surplus is the sum of areas DS1, DS2 and
SS2, while producers' surplus is area SS1. Thus, these interventions while setting price have
undesirable effect of reducing the global welfare by an amount equal to the sum of areas DW1 and
DW2. The amount equivalent to this area is called as the deadweight loss. In general, regulated tariff
is the major source of deadweight loss. From electric market point of view, the network constraints
can be a major source of creation of deadweight loss. This will be discussed in further chapters.

SHORT-RUN AND LONG-RUN

In section 2.2, we have seen the consumer behavior and associated cost functions. Recall production
function of Figure 2.5. Equation 2.2 suggests that the output is a function of only a single factor of
production x'. In reality, the output is a function of many factors of production. Thus,

y=f(x1,x1..............xn)............................................................................................(2.6)

However, all the production factors do not affect the output in the same way. Some factors make
immediate changes while others take long time to be effective. In other words, some factors affect the
production in the short-run while the others do so in the long-run. There is no clear-cut distinction
about the time horizon which divides the production factors between short-run and long-run. In short-
run, some of the production factors are fixed. Long-run window provides the time span sufficient
enough for various factors to be adjusted. To provide an analogy, a reactive power support by means
of a capacitor bank can be termed as a short-run factor as compared to the construction of an entirely
new parallel circuit for increasing capacity of a heavily loaded transmission corridor.

In the short-run, the output often depends on a single production factor while the rest of the factors
are considered to be fixed. Thus, cost functions and marginal cost functions depicted in Figure 2.8 are
essentially short-run cost function and short-run marginal costs function.

The short-run behavior of a commodity producing firm in a perfectly competitive market can be
explained as follows. In this type of market, the only option left to the firm is to maximize its profit is
by adjusting its output, as it can not influence the market price. This can be posed as follows:

...............................................................................................(2.7)

Optimum will be achieved when

.....................................................................................................(2.8)
Thus, to maximize its profit, the firm will raise its production up to that level at which the marginal
cost is equal to the market price. This point will be elaborated further in section 2.8 where we will see
the profit maximization of a firm in a perfectly competitive market.

In the long-run, the firm will have more degrees of freedom to work with. Thus, defining a long-run
cost function is a complex task. The long-run cost function as against the short-run cost function will
be given as in equation 2.6. It should be noted that the long-run cost function (Cost LR) is the solution
of an optimization problem.

such that
y = f(x1,x2,........xn)

wi is the unit cost of production of factor xi. It is worthwhile to note that in general, the decisions
about power system operation are short-run, while those associated with planning are the long-run
solutions. For example, optimal scheduling of generation requires generation costs to be known. These
costs are short-run costs that assume the plant size and the network capacity to be fixed.

VARIOUS COSTS OF PRODUCTION

Let us now see what the components of production cost are for a firm. In the short run, some of the
factors of production are fixed. The cost associated with these factors does not depend on the amount
produced and is thus a fixed cost. Let us take an example of a generating company. For a certain fixed
maximum capacity of a plant, the cost of land and associated machinery does not depend on the
output of the plant. On the other hand, the quantity of fuel consumed by this plant depends on the
output. Thus, the cost of fuel is a variable cost.

The cost function for a generating plant can be defined as function of level of MW output. It can be
given as follows:

.......................................................................................................(2.9)

Where, is the variable cost and the fixed cost. The average cost is defined as the cost per MW output
of a plant. It is the sum of average fixed cost and average variable cost. Thus, the average cost is
given as:

..................................................................................(2.10)

RELATIONSHIP BETWEEN LONG RUN AND SHORT RUN AVERAGE COSTS

As discussed earlier, there are many factors of production that are fixed in the short-run but variable
in the long-run. That is why a firm's short-run cost curve differs from that of a long-run cost curve.
For a firm, to increase its output in short-run requires increase in those factors of production which
show their effect in short-run. For example, for a vehicle manufacturer, increasing number of workers
is one such solution in order to increase number of vehicles produced. In the long-run, it can increase
its vehicle production by building more number of factories. In short-run calculation, the costs of these
factories will not figure and are supposed to be fixed, while in the long-run, they are variable. Figure
2.14 shows three average short run cost curves for three sizes of vehicle manufacturing factories
small, medium and large. The figure also shows the long-run average cost curve. As the firm moves
along the long-run cost curve, it adjusts the factory size to the quantity of production.

As seen from the figure, long run average cost curve has a much flatter U shaped curve than short run
average cost curve. All short-run curves lie on or above the long run cost curve. The long-run cost
curve provides a lower envelope for all short run average cost curves. This peculiar relationship exists
because of flexibility associated while obtaining long-run cost curve. In short, the firm gets to choose
which short run curve it wishes to employ.
 Figure 2.14: Long run AC and Short run AC

As the firm moves from a small size factory to medium size factory, the cost per vehicle starts
reducing till a certain point. As long as the average production cost decreases, the product is said to
display economies of scale. In the electricity business, the transmission systems are said to exhibit
economies of scale.

PERFECTLY COMPETITIVE MARKET

A perfectly competitive market, in brief terms, can be explained as a market form in which no
consumer or producer can influence the market price. Perfect competition has a market equilibrium
which is Pareto efficient. The defining characteristics of a perfectly competitive market can be given as
follows:

1. Atomicity : An atomic market is the one in which there are a large number of small producers
and consumers They are so small that individual actions have no significant impact on others.
Firms are price takers.
2. Complete information : All consumers and sellers know the prices set by all firms.
3. Homogeneity : All firms sell an identical product.
4. Free entry : No firm has barrier for entry into or exit out of market.
5. Uniform price: Each firm charges the same price in the market.

In addition to this, there are obvious behavioral assumptions such that the consumers try to maximize
utility while suppliers try to maximize profit.

Markets for agricultural products with sufficiently large number of suppliers and consumers can be
assumed to be a perfectly competitive market. On the other hand, the electricity markets seldom
exhibit prefect competition due to less number of dominant players.

A firm in a competitive market, like most other firms, tries to maximize profit, which equals total
revenue minus total cost. We have seen the concepts of total, average and marginal costs of a firm.
Similar concepts pertaining to revenue of a firm exist. Total revenue (TR) is price times the quantity (
*,q), and average revenue (AR) is total revenue (*,q ) divided by the quantity (q). Therefore, for all
firms, average revenue equals the price of the good. On the other hand, marginal revenue (MR) is the
change in total revenue from the sale of each additional output. Total revenue is (*,q ) and is fixed
for a competitive firm (firm being a price taker in perfectly competitive market). Therefore, when q
rises by 1 unit, total revenue rises by rupees. Therefore, for firms in perfectly competitive markets,
marginal revenue equals the price of the good.

The Firm's Supply Decision under Perfect Competition

Equation 2.8 showed that in order to maximize the profit, the firm will raise its production up to that
level at which the marginal cost is equal to market price. Let us explain this point graphically with the
help of the concept of marginal revenue. To understand firm's supply decision under perfect
competition, refer to Figure 2.15.

Figure 2.15: Profit maximization for a firm under perfect competition

The marginal-cost curve (MC) is upward sloping. The average cost curve (AC) is typically U shaped.
And the marginal cost curve crosses the average cost curve at the minimum of average cost. The
figure also shows a horizontal line at the market price ( *). The price line is horizontal because the
firm is a price taker. The price of the firm's output is the same, regardless of the quantity that the firm
decides to produce. We have mentioned that for a competitive firm, the firm's price equals both its
average revenue (AR) and its marginal revenue (MR).

We can use Figure 2.15 to find the quantity of output that maximizes profit. Imagine that the firm is
producing at q1. At this level of output, marginal revenue is greater than marginal cost. That is, if the
firm raised its level of production and sales by 1 unit, the additional revenue (MR1) would exceed the
additional costs (MC1). The profit, which equals total revenue minus total cost, would increase. Hence,
if marginal revenue is greater than marginal cost, as it is at q1, the firm can increase profit by
increasing production.
Now suppose the firm produces quantity q2. In this case, marginal cost is greater than marginal
revenue. If the firm reduced production by 1 unit, the costs saved (MC2) would exceed the revenue
lost (MR2). Therefore, if marginal revenue is less than marginal cost, as it is at q2, the firm can
increase profit by reducing production.

Where do these marginal adjustments to the level of production end? Regardless of whether the firm
begins with production at a low level (such as q1) or at a high level (such as q2), the firm will
eventually adjust production until the quantity produced reaches q*. Thus, a generic rule for profit
maximization can be laid down as follows: At the profit-maximizing level of output, marginal revenue
and marginal cost are exactly equal.

We can now see how a competitive firm decides the quantity of its good to supply the market.
Because a competitive firm is a price taker, its marginal revenue equals the market price. For any
given price, the output that maximizes the profit of the competitive firm is found by looking at the
intersection of the price with the marginal cost curve. However, this analysis does not hold good if the
market is not perfectly competitive. There are various forms of imperfect competition and then the
firms are not mere price takers. The firms can strategically decide upon their actions that reflect in the
market price.
 Module 3: Philosophy of Market Models
Creating trading arrangements to exchange a commodity called electrical energy adds complexity to
the overall process of generating and transferring the same to the end customer. Deviating from a
vertically integrated structure towards unbundled operation requires structural and constitutional
changes. This is true for any commodity market that is undergoing the paradigm change. In this
module, we discuss various market models for trading of electricity and the need for designing the
same. Eventually, we answer the following important questions:

What are the possible ways in which buyers and sellers can trade electrical energy?
Which section of consumers has a choice of selecting their energy provider?
What are the peculiarities of electricity that make market arrangements of this commodity different
from other commodities?
What is the role and involvement of system operator in the market decisions?

Philosophy of market models is influenced by the multiple ways in which the above questions can be
answered. In this module, first we explain choices of industry structures for trading of electricity.
Next, we present the peculiarities associated with electricity and discuss how these characteristics
influence the market design. The peculiar characteristics of electricity lead to evolvement of various
markets in which the same electricity is treated as a different product. The arrangement of these
markets and their alignment and linkage with main energy market is an issue of market architecture.
The market architecture intricacies are discussed in detail.

MARKET MODELS BASED ON CONTRACTUAL ARRANGEMENTS

As mentioned earlier, unbundling of the conventional vertically integrated power system creates
groups of various commercial and technical activities. Since one of the major aims of deregulation is
introduction of competition, it is worthwhile to explore every avenue where competition can be
introduced. Eventually, competition provides a choice for entities to choose another entity or a group
of entities to do a profitable transaction. In electricity parlance, either the load or an entity
representing a group of loads gets a choice to select its energy provider, or there may exist some
mechanism which would cater to the electrical energy needs of these loads at a competitive level.

The former mechanism essentially requires bilateral involvement of the entities who wish to get into a
power buy and sell contract. In this, the sellers and buyers mutually agree upon the terms and
conditions, including the price and time of delivery. A repetitive bilateral interaction between buyers
and sellers may lead to an equilibrium point where everyone is happy. Alternatively, a similar result
would be obtained if a common exchange for the commodity is set up, where, buyers and sellers,
instead of interacting with each other, communicate their expectations to this marketplace. This
represents a simultaneous market clearing process and a common market price of electrical
commodity.

While moving from a vertically integrated structure to a competitive one, various policy and structural
issues crop up. One of the important concerns is regarding the entity that should be allowed to take
part in competitive activity. Similarly, issue of rearrangement of various elements of power system,
when a new set of rules is introduced to buy and sell power, also needs to be addressed. It is obvious
that the commercial arrangements and virtual boundaries between various functional entities can take
many shapes and forms. Consequently, various models can be classified according to the levels at
which the entities are given the choice of buying or selling electricity.

Various trading models can be proposed based on the above discussion. The choice of choosing a
model is a policy decision and is dominated by various prevailing conditions. They need to be
accounted for before making structural changes. In [1], four basic models of industry structure are
suggested. These are:

1. Monopoly model
2. Single buyer mode
3. Wholesale competition model
4. Retail competition model

Every model needs different amount of structural change and rearrangements of functions in the
industry. These models are discussed next.

Monopoly Model

In this model, a single entity takes care of all the businesses such as generation, transmission and
distribution of electric power to the end users. One of the versions of this model is shown in Figure
3.1(A). In this, a single utility integrates the generation, transmission and distribution of electricity.
Usually (but not necessarily), in this kind of model, the monopoly lies with the Government. It is quite
natural that this kind of model should have strict regulation in order to protect end consumers against
monopoly. Most of the electric power systems followed this model prior to deregulation.

Another version of the monopoly model is shown in Figure 3.1(B). In this model, generation and
transmission are integrated and operated by a single utility and it sells the energy to local distribution
companies, which themselves represent local monopolies.
 Figure 3.1: Two different versions of monopoly model
Single Buyer Model

In this model, as shown in Figure 3.2, there is competition in the wholesale sector, i.e., generation.
Here, the single buyer agency buys power from Independent Power Producers (IPPs) in addition to its
own generation. The power purchasing agency in turn sells it to state distribution utilities or
distribution companies in the service area. All power generated by generating companies (Gencos)
must be sold only to a purchasing agency and not to any other agency. Distribution companies
(Discoms) are only able to purchase from the single buyer agency. They do not have a choice of
choosing their power supplier.

In this model, sales from power pool to retailers take place at a pre-set tariff price. The single buyer
or the existing utility makes a long term contract with IPPs. A contract is necessary because, without
it, a generator would be reluctant to invest large amounts of capital in a generating plant. The
contracts are generally of life-of-plant type, indicating sale of all capacity of generating units for its
lifetime.

Figure 3.3 shows another version of this model, which has further evolved from the original single
buyer model. In this model, the single buyer does not own any generation and buys all the power
from IPPs. The distribution and retail activities are also disaggregated. This model has an advantage of
introducing some competition between generators without the expense of setting up a competitive
market. The tariff set by the purchasing agency must be regulated because it has monopoly over the
Discos while monopsony over the IPPs. The single buyer model is looked upon as a way of attracting
private participation in the generation sector, especially in the developing countries.

In this model, transmission and distribution network can be owned and operated by State and
Regional transmission utilities. Inter-state tie line should be sufficient to maintain a loose regional
power pool. Merits and demerits of this model are as follows:

Merits:

Private participation in power generation

Introduction of some competition without expensive set up for a competitive market.

Figure 3.2: Single buyer model

Demerits:

Long term contracts. Setting up a contract is problematic.

No true competition.
Price is not decided by demand-supply interaction.
End consumers' price is regulated .

Wholesale Competition Model

This model is one step closer towards competition. There is an organized market in which the
generators can sell their energy at competitive rates. The market may be organized either by a
separate entity or may be run by the system operator itself. There is not much choice for the end
user. The end user is still affiliated to the Discom or retailer working in that geographical area of
operation. The large customers or the bulk customers, so to say, are privileged to choose their energy
provider. However, the definition of bulk customer is a subjective matter and changes from system to
system.
 Figure 3.3: Single buyer model with only IPPs

This model, as shown in Figure 3.4, provides the choice of supplier to Discoms, along with competition
in generation. Implementation of this model requires open access to the transmission network. Also, a
wholesale spot market needs to be developed. Since this model permits open access to the
transmission wires, it gives the IPPs to choose an alternative buyer. Discoms can purchase energy for
their customers either from a wholesale market or through long term contracts with generators.

The customers within a service area still have no choice of supplier. They will be served by a Discom in
their area. With this model, the Discoms are under Universal Service Obligation (USO), as they have
monopoly over the customers. They own and operate the distribution wires. The transmission network
is owned and maintained either by government and/or private transmission companies. System
operators manage the centrally accomplished task of operation and control.
 Figure 3.4: Wholesale Competition Model

The model provides a competitive environment for generators because the wholesale price is
determined by the interaction between supply and demand. In contrast, the retail price of electrical
energy remains regulated because the small consumers still do not have a choice for their supplier.
The distribution companies are then exposed to vagaries of the wholesale price of the commodity. The
merits and demerits of this model are as follows:

Merits:

Choice of seller provided for Discoms and bulk consumers.

The buyers and sellers can make forward contracts or buy from a wholesale marketplace.
The price is decided by interaction between demand and supply. Hence, indicates truly
competitive price.

Demerits:

The end consumer still doesn't have a choice. It buys power from the affiliated Discom.
Rates for end consumers are regulated rather than competitive.
Discoms face competition at wholesale level, while their returns are regulated.
Structural and institutional changes required at wholesale level.

Retail Competition Model

In this model, as shown in Figure 3.5, all customers have access to competing generators either
directly or through their choice of retailer. This would have complete separation of both generation
and retailing from the transport business at both transmission and distribution levels. Both,
transmission and distribution wires provide open access in this model. There would also be free entry
for retailers. In this model, retailing is a function that does not require the ownership of distribution
wires, although, the owner of distribution wires can also compete as a retailer.

This model is a multi-buyer, multi-seller model and the power pool in this model acts like an
auctioneer. It behaves like a single transporter, moving power to facilitate bilateral trading and this is
achieved through an integrated network of wires. In this pooling arrangement, there is a provision for
bidding into a spot market to facilitate merit order dispatch. The pool matches the supply and demand
and determines the spot price for each hour of the day. It collects money from purchasers and
distributes it to producers.
The advantage of this model over monopoly utilities is that competition is introduced in both
wholesale and retail areas of the system. This model is supposed to be a truly deregulated power
market model. The retail price is no longer regulated because small consumers can change their
retailer for better price options. This model is economically efficient as the price is set by
interaction of demand and supply. In wholesale competition model, with relatively few
customers, all of them regulated Discoms, a spot market can be preferable but not essential.

However, in retail competition model, spot markets become essential, since contractual arrangements
between customers and producers are carried out over a network owned by a third party. In retail
competition model, metering becomes a major problem. If the number of customers are increasing
and metering capability for all the customers is not sufficient, it may create logistical problem and
provoke disputes.
 Figure 3.5: Retail Competition Model

Merits:

Supposed to be 100% deregulated model.

Every consumer has a choice of buying power.
The price is decided by interaction of demand and supply. Hence, it is truly competitive price.
There is no regulation in energy pricing.

Demerits:
Need constitutional and structural changes at both, wholesale and retail level.
Extremely complex settlement system due to large number of participants.
Requirement of additional infrastructural support.

Comparison of Various Market Model

A comparison between the four models discussed above is provided in Table 3.1. The
attributes for comparison are chosen in such a manner that the difference between two truly
competitive models - wholesale competition and retail competition gets significance. It is
then obvious that the monopoly and single buyer models will emerge to be similar models
when compared on the chosen attributes.
 Single Wholesale Retail
Attribute Monopoly
Buyer Competition Competition

Degree of
0 1 2 3
deregulation*

Number of
1 1 Many Many
buyers

Number of
1 Many Many Many
seller

Everbody
Choice Discoms and including
No No
available big customers small
consumers

Requirement Preferable,
of spot No No but not Essential
market essential

Transmission
Transmission as well as
Open access NA NA
network distribution
network

Regulated
Small
price to be Everbody Everbody None
consumers
paid by

ELECTRICITY VIS-A-VIS OTHER COMMODITIES

The classification of market models based on contractual agreements discussed in the previous section
can be applied to most of the commodities that are traded in the market, if we assume a certain level
of abstraction by presenting only the buyers and sellers. However, when it comes to electricity' as a
commodity, the same laws of economics or commercial trade arrangements may not hold good. This
is because, electricity as a commodity bears different characteristics from other commodities, or
rather, electricity is physically different from other commodities. This fact complicates the procedure
of electricity trading. In other words, the trade is not as simple as an interaction between two entities:
buyer and seller. The interdependencies of actions taken by various participants (primarily generators
and loads), mandate somebody to takeover the control of real time activities. This somebody is the
system operator, who makes sure that the whole system runs reliably and thus kept in synchronism.
Thus, it is worthwhile to understand the distinguishing features of electricity as a commodity, which
are presented next.

Distinguishing Features of Electricity as a Commodity

There are three basic distinguishing features of electricity. These are associated with electricity due to
its physical nature. These three basic features effectively lead to one distinguishing feature of this
commodity, the one that has commercial implications. Let us see these in details

Real Time Demand Supply Balance

Electricity can not be stored in bulk. Other commodities can be manufactured and kept in a warehouse
until the demand for the same is sensed. A manufacturer of other commodities gets sufficient
flexibility in planning the manufacturing activity and coordinating the dispatch. The same is not true
for electricity. The demand for electricity needs to be satisfied on real time basis.

The parties involved in electricity trade perhaps would like to do it through forward contracts . These
can be contracts for physical delivery or financial in nature. In many power markets, bulk trade of
electricity (> 80%) is done through forward contracts. Forward contracts can be done years ahead.
When a certain amount of electricity is bought in the forward contract, it is the estimate of the buyer,
how much it is likely to consume during actual delivery time. However, in real time, the actual
consumption may not match the predicted consumption that had been forecasted at the time of doing
forward trade. This difference is called as imbalance. Knowledge about this imbalance is exposed only
during real time operation or slightly before that. In this case, the system operator or some other
market mechanism stands ready to make up the imbalances (either on positive or negative side).

Due to storage limitation, the supply-demand matching decision needs to be done on a competitive
basis by letting supply and demand interact with each other. The operator buys and sells these
imbalances through some commercial mechanism. Due to this feature of electricity, an issue related to
the speed of operation pitches in. The system operator, while making a provision for imbalances, has
to take into consideration various network interdependencies. The system operator always has to
communicate with the active participants to tell them which generators should increase their output
and which ones should decrease it. This activity is called scheduling in advance and dispatch in real
time. Since the system operator has to work with seconds to spare, a delivery system to make up for
imbalances has to be in place. In real time, the only time available with system operator is what is
allowed by the energy stored in rotating masses of huge interconnected grid.

Thus, this exceptional feature of electricity leads to two issues related to power market design:
Imbalances and Scheduling and Dispatch. The question is how these difficult tasks get reflected in the
rules of marketplaces.

Power Flows Obey Laws of Physics

The electric power can not be told as to where and how it should travel, once the injection and take-
off points are decided. The electric power flow over transmission lines obey laws of physics.
Effectively, electric power can not be stopped from flowing on a transmission line that is already
hitting its power carrying capacity. The system operator has to ensure that none of the lines get
overloaded. To do this, only freedom left with it is the selection of pattern of nodal injections (either
generation or load).

Thus, any arbitrary set of forward contracts can not be scheduled by the system operator as this may
lead to exceeding of limits of physical parameters of some of the power system elements. Allowing
only the practically feasible set of transactions during scheduling and further making corrections while
dispatching so as to keep line loadings within limits is usually termed as congestion management.

The concept of network congestion is shown by a simple lossless system in Figure 3.6. In this,
generator A is a cheaper generator than generator B and hence, it gets a contract of satisfying the
demand of load at bus 3 by generating 18 MW. The dispatch would be as shown in Figure 3.6(A). The
power flow over all lines would be dictated by the reactance of parallel paths. In this case, let us
assume that reactance of all three lines are same. Thus, two parallel paths are provided so as to
transfer power of generator A to load at bus 3, with ratio of reactance 2:1. Obviously, the power will
flow in opposite ratio on these paths. The flows are shown in Figure 3.6(A). However, if the physical
properties of the line connecting nodes 1 and 2 state that it can carry only 3 MW, then the dispatch
shown in Figure 3.6(A) left hand side is not practically feasible. To correct it, generator B is asked to
generate 4.5 MW and generator A is asked to step down by 4.5 MW, leading to dispatch shown in
Figure 3.6(B). This rearrangement of nodal injections is one of the means of congestion management,
which is peculiar to electricity. We will discuss more about this in a separate module on congestion
management.

Figure 3.6: Concept of network congestion

Generator Product Compatibility and Interactions

To ensure reliable delivery of electricity, only generation by generators at injection points and take-off
by loads at take-off points is not sufficient. The system operator must make arrangements for
provision of allied services necessary to do this. These allied services are usually referred to as the
ancillary services. Provision of reactive power, operating reserves are some of the commonly required
ancillary services. Mostly, ancillary services are provided by generators. In this case, one is likely to
witness the interdependencies involved in providing these services. In other words, the production of
ancillary services is also dependent on production of energy. Then, the same generator is said to be
providing two different products: energy and ancillary services.

This complicates the matter because the single generator can be simultaneously needed to produce
multiple outputs, or to produce ancillary service rather than energy. This complication is shown in
Figure 3.7, where, a generator's capacity is divided into various products. The defining question is how
much of capacity should be allocated to each product? In centralized markets (explained later), the
system operator does a joint optimization, taking into account various technical and commercial
parameters of a generator to allocate it's full capacity to each of the products. module 6 is devoted to
ancillary service management where these issues will be discussed more elaborately.
 Figure 3.7: Generation capacity allocation to various products

Unusual Price Variation

The combined effect of various peculiarities of electricity is that it has large temporal variation in its
price. It is not prudent to run all generators throughout the day. Rather, the most economical
generators can be run throughout the day. Effectively, the price of electricity will be low during low
demand period. However, during peak demand situation, the costly generators are brought on-line
and the price of electricity goes high. Thus, marginal cost of producing energy will vary throughout the
day. Such rapid cyclic variations in the price of a commodity are unusual, and arise due to peculiarities
associated with electricity, basically, the characteristic of matching supply and demand on real time
basis. It should be noted that this peculiarity of electricity has arrived because of one of the basic
physical properties associated with it.

Effects of Peculiarity: Four Pillars of Market Design

We have seen the characteristic features of electricity when compared with other commodities. How
do these affect the trading activities of this commodity? For example, what if network congestion does
not allow a set of transactions to be feasible? Should the generator sell its generation capability in a
single market that makes provision for energy as well as reserves, or should there be different
markets for the same? Some subtle questions like these provide food for thought when designing
criteria of markets are to be determined.

Hunt in [1] has described the design issues arising out of characteristics of electricity as pillars of
market design. These are:

Imbalance
Scheduling and Dispatch
Congestion Management
Ancillary Services

Figure 3.8 shows four pillars of market design arising due to the basic characteristics of electricity.
 Figure 3.8: Four Pillars of Market Design

The design of market revolves around the four pillars described above. It also depends on how and
where these issues are accommodated in the whole process of market mechanism. Some of the pillars
lead to creation of separate markets. Eventually, this gives rise to the issue of market architecture,
which is nothing but arrangement and classification of these markets. Finally, these markets can be
integrated into one efficient market or there can be cascaded markets. The architectural aspects of
market design are discussed next.

MARKET ARCHITECTURE
Stoft in [2] defines market architecture as a map of its component submarkets. This map includes the
type of each market and the linkage between them. Where does this concept of multiple markets
come from? The answer can be traced back to various peculiarities associated with electricity. Four
pillars of market design tend to cast the same electric energy into various products, which are
characterized by separate individual markets. Moreover, there are various modes of energy contracts
depending upon when energy trades are done. This again gives rise to market mechanisms based on
timeline of trading.

The submarkets of a power market include the wholesale spot market, wholesale forward markets and
markets for ancillary services. Somewhere in between is embedded the market for transmission
capacity. This can be a separate market altogether or can be integrated with the energy market that
takes place near real time. Similar is the case with ancillary service market. The best way to
categorize alternative trading models is on the degree to which operational arrangements and
commercial arrangements for scheduling, imbalances, congestion and ancillary services are integrated
with spot markets. Two models are most common: integrated or centralized and decentralized.

In the rest of the module, we will give more stress on how various markets for energy and other
products are organized. For the sake of understanding, we will not go into the intricacies involved in
various modes of arrangement and levels of competition discussed in Section 3.2. We will just
represent market by a set of sellers and buyers. For this, we do the abstraction of the market as
shown in Figure 3.9, indicating sellers and buyers with some interaction facilitator. This abstraction
gets rid of questions about ownership of transmission network, power exchange, distribution network,
as well as doesn't bother about whom the buyer represents or buys for its own. The relevant details
about the same will be discussed at appropriate places. First, let us see how markets for energy are
arranged.

Timeline for Various Energy Markets

There are many ways depending on the time of hand-shaking, where buyers and sellers can do the
transaction. Figure 3.10 shows various modes of trading based on the time-line.
Following are the common modes in which the electric energy can be traded:

1. Bilateral contracts
2. Spot market

a. Day ahead markets (Power Exchange or through pool)

b. Real time market (through pool)

Figure 3.9: Abstraction of market concept

Trading for power delivered in any particular minute begins years in advance and continues until real
time, the actual time at which the power flows out of a generator and into a load. This is accomplished
by a sequence of overlapping markets. The earliest amongst these are forward markets that trade
non-standard, long term, bilateral contracts. This generally represents energy trading between buyers
and sellers directly for the mutually agreed price. This type of trading stops about one day prior to real
time. At that point, the day-ahead market is held. The day-ahead market is often followed by a real-
time market.

The term - spot market is used with different interpretations associated with it. According to
definitions in some of the systems, the spot market includes day-ahead and real-time market, while in
others; it just includes the real-time market. Similarly, drawing line between spot and forward
markets is not clear. According to one definition, all the markets before the real-time market can be
classified as forward markets. This is because, in many forward markets, including day-ahead market,
traders need not own a generator to sell power. If power is not delivered in real time, then the
supplier must purchase replacement power at the real time rate and fulfill the contract. A customer
who buys power in a forward market will receive either electricity delivered by the seller or a financial
compensation. Any power that is sold in the day-ahead market, but not delivered in real time, is
deemed to be purchased in real time at the real time price of energy. The combination of day-ahead
and real-time market is popularly known as multi-settlement market system in USA.

Another way to distinguish between forward and spot markets is by considering day-ahead and real-
time markets as spot markets, while all trades taking place before that are termed as forward or
bilateral trades. This segregation emerges because both, the day-ahead as well as real-time markets
provide a system price which holds for all the market trades done through it. On the other hand, in
bilateral or forward trades, there is no single market price as such. In the rest of the module, we
prefer to define the spot market as defined just above.

There is little doubt about what should be the nature of settlements based on timeline. Much ahead of
real time, i.e., more than a week, month or years ahead, bilateral contracts provide the best manner
of trading power. One is very unlikely to have bilateral contracts near real time. The reason is that the
settlements of bilateral contracts take place very slowly. Near real time, it is prudent to have a
centrally organized market as the security and reliability issues can be tackled centrally rather than
bilaterally. Even, the day-ahead market can be centrally organized. It can take the form of power
exchange or the pool. In other words, day-ahead market can be organized by a separate entity or it
can be integrated with the system operator activities. If the latter is adopted, it is popularly known as
a pool structure.

In general, real time transactions require central coordination, while week-ahead trades do not require
the same. Somewhere in between are dividing lines that describe the system operator's diminishing
role in forward markets. Where to draw those lines is the central controversy of power market design.
A larger role for the system operator implies a smaller role for private, profit making entities.

Bilateral/Forward Contracts

Bilateral trading generally involves two parties interacting with each other: a buyer and a seller. The
characteristic of bilateral trades is that the price of a transaction is set independently by the parties
involved. There is no market clearing price as such. Since, electricity can not be stored, it creates a
wide fluctuation in the spot price. Forward contracts provide generators and loads with a means of
hedging their exposure to fluctuations in the spot price of electricity. The generators can negotiate a
price for their output prior to the moment of producing it. Similarly, properly structured forward
contracts provide buyers with the ability to lock in a fixed price for a fixed quantity of electricity well in
advance of delivery and consumption. Indeed, if a buyer's actual energy usage matches its forward
market purchases, it can achieve a benefit of complete price certainty in the face of real time price
volatility.
 Figure 3.10: Seller buyer interaction based on timeline

Depending upon the quantity of power and time, the buyers and sellers resort to different forms of
trading:

Long Term Contracts: This type of trade generally includes contract for a large amount of
power for a long time period. These types of contracts are negotiated privately and the terms
and conditions are such that they suit both the parties involved in the transaction.
Trading Over The Counter: These transactions involve smaller amounts of energy to be
delivered. For example, the amount of energy to be delivered during different periods of the
hour, day, etc. This type of trading has much lower transaction costs and is used by producers
and consumers to refine their positions before real time.
Electronic Trading: In this, participants can enter offers to buy energy and bids to sell
energy directly in a computerized marketplace. The participants can observe the quantities
and offers/ bids submitted by all participants, but do not know the party involved. The
software in the exchange couples the matching offers. It checks whether for a newly entered
bid, if there is matching offer whose price is greater than or equal to price of the bid. If no
match is found, the bid is added to the list of outstanding bids until a new offer matches it.
Otherwise, it lapses after the market is closed. The same process is repeated after a new bid is
 entered. There is no market clearing price as such.

The Spot Market

As we have discussed in module 2, a market for any commodity provides an environment for buyers
and sellers to interact and agree on transactions, generally, the quantity and price. These interactions
progressively lead to an equilibrium point at which the price clears the market, that is, the supply is
equal to demand. If electrical energy is to be traded according to a mechanism in which the buyers
and sellers are free to interact individually, the equilibrium between the production and the
consumption can be set through repetitive interaction. In this scheme of attaining equilibrium, the
consumers make an estimation of their consumption before entering into a contract. The generators
schedule the production of their units to deliver at the agreed time the energy that they have agreed
to sell. However, in practice, neither party can meet its contractual obligation with perfect accuracy
because, for example, from a load's point of view, the actual demand of a group of customers is never
exactly equal to the value forecasted. Changes in weather and due to some other externalities, the
day ahead or before real time estimation of load consumption can have deviation from that done few
months or years back, while doing the contract. Also, unforeseen problems may prevent generating
units from delivering the contracted amount of energy.

It can be concluded that, while a large proportion of the electrical energy can be traded through an
unmanaged open market in terms of forward contracts, such a market may not necessarily lead to an
equilibrium that replicates real time scenario. Thus, an intermediate stage is necessary, where a
managed spot market can provide a mechanism for balancing load and generation. This market should
supersede the open energy market as the time of delivery approaches. Its function is to match
residual load and generation by adjusting the production of flexible generators and curtailing the
demand of willing customers.

In many real life markets, more than 80% of the energy traded is through the forward or bilateral
contracts. The rest is traded through the spot market. In a multi-settlement market (typically
practiced in some of the markets in USA), the spot market is sometimes made of two markets: Day
Ahead (DA) market and a Real Time (RT) market. The DA market is run for each hour or half hour of
the next day. The RT market is always run by a system operator, while the day-ahead market may or
may not be run by the system operator. In both cases, the general principle of market clearing is the
same. This and other related issues are discussed next.

Spot Market Clearing

For the sake of understanding, let us assume that the market is run by an entity called Power
Exchange (PX). The power exchange operates much like a stock exchange. The buyers and sellers
enter their needs into the power exchange. For example, a buyer would say, I need up to 20 MW
between 1600 hours and1700 hours IST. I would pay INR 3.5/ kWhr, whereas, the seller would enter
his demand as, I have 100 MW and would like to sell it at INR 4/ kWhr.

When they transact with the power exchange, buyers and sellers are really talking to the marketplace
and not the individual buyers and sellers. As in a stock exchange, the power exchange constantly
updates and posts a market clearing price (MCP), which is the current price at which the transactions
are being done. Note that when buyers and sellers communicate with the power exchange, they dont
know whom they are dealing with. The general step by step process of settling this market is as
follows:

1. Generating companies submit bids to supply a certain amount of electrical energy at a certain
price for the period under consideration. Usually, the period is an hour or half an hour. The
bids are ranked in order of increasing price. From this, a curve that shows bid price as a
function of bid quantity is built, which is commonly known as supply curve. Supply curve is a
plot with price on y axis and quantity on x axis.
 2. Similarly, demand curve is established by asking consumers to submit offers specifying
quantity and price and ranking these offers in decreasing order of price. If the load is willing to
adjust its consumption with price, the load is said to have demand elasticity. If the load is
firm, the demand curve will take the form of a vertical line with x axis intersection indicating
total cumulative firm demand.
3. The intersection of supply and demand curves represents the market equilibrium. At this
point, the supply matches the demand. This price is known as Market Clearing Price (MCP) or
System Marginal Price (SMP). All the bids submitted at a price lower than or equal to the
market clearing price are accepted and the generators are scheduled for that much amount of
power for that particular time period under consideration. Similarly, all the offers submitted at
a price greater than or equal to the market clearing price are accepted.
4. As for settlement, the generators are paid this MCP for every MWh they are scheduled for,
while loads pay the MCP for every MWh they are cleared for.

Illustrative Example for PX Clearing

Suppose there is a central power exchange in which all players in the market send bids and offers.
Table 3.2 shows the offers and bids supplied to the central power exchange for a particular hour of the
next day, say 10:00 AM to 11:00 AM.

Once the buyers and sellers provide offers and bids, the power exchange forms an aggregate supply
curve and aggregate demand curve. The curves are plotted on the coordinates of supply (and
demand) and price as shown in the Figure 3.11. The point of intersection of the two curves determines
the market-clearing price (MCP). At this point, the supply satisfies the demand.

From the intersection of supply and demand curves, the MCP would be set to 3200 INR/MWh and 450
MWh will be traded through the central power exchange. The MCP is the price of electric energy that is
paid by consumers trading through the power exchange. The sellers are also paid at a price equal to
the MCP. MCP is the highest sell bid or lowest buy bid accepted in the auction. The generator S2' is
called the marginal generator as its bid sets the MCP.

Over and above the forward contracts, the participants trade the residuals through the power
exchange. The objective of the power exchange clearing is to maximize the social welfare as explained
in the earlier module. It is the sum of generator surplus and the load surplus.

Quantity Price
Company
(MW) (INR)

S1 200 2400

S1 50 3000

S1 50 4000

Bids S2 150 3200

S2 50 3400

S3 100 2600

S3 50 3600
 D1 50 2600

D1 100 4600

D2 50 2200
offers
D2 150 4400

D3 50 2000

D3 200 5000

Table 3.2: Bids and offers in the power exchange

Discriminatory or Non-discriminatory Pricing?

There are few questions which are likely to remain unanswered regarding the settlement procedure
adopted in the above market clearing process. One is likely to get surprised to see all of the
generators being paid the MCP, rather than at individual bid. Except the marginal generator, all other
generators are willing to produce power for lesser price than the MCP. Then, why are they not paid
their asking price? Paying them their asking price would have reduced the average price of electricity.

Paying generators as per their asking price is known as pay-as-bid scheme. The main reason why pay-
as-bid scheme is not adopted is that it would discourage generators from submitting bids that reflect
their marginal cost of production. Basically, the notion that the average price of electricity would
decrease by adopting pay-as-bid scheme is based on the assumption that the generators would
continue to bid in the same way as they do in the marginal pricing scheme. However, this is not true.
All the generators would instead try to guess what the MCP is likely to be and would bid at that level
to collect the maximum revenue. While doing so, some low cost generators would bid too high. Then,
in the market clearing process, these generators would not get selected and be replaced by some
other generators that have higher marginal cost of production. The MCP would then be somewhat
higher than it ought to be. Furthermore, this substitution is economically inefficient because optimal
use is not made of the available resources. In addition, the generators are likely to increase their
prices slightly to compensate themselves for additional risk of losing revenue because of uncertainty of
MCP. An attempt to reduce the price of electricity would therefore result in a price increase.

On the other hand, in marginal pricing scheme, a seller is certain that it will be paid no less than its
cost of production if he bids at marginal cost, and may be paid more. If a seller bids less than his
marginal cost, it would lose money because his bid may set the MCP. If it bids more than its marginal
cost, it may bid more than other sellers and fail to be selected in the auction. If the seller's bid sets
the MCP, then it would recover it's running cost and if the MCP is higher than it's marginal cost, then it
would earn profit or contribution to fixed cost. It is worthwhile to know that the supply curve, being a
derivative of cost function, does not consider the fixed costs.
 Figure 3.11: Calculation of market clearing price (multiple price by 200)

Simple Bids or Complex Bids?

In the illustrative example provided above, the generators submitted simple bids consisting of price-
quantity pairs. In some system operator run markets, (typically known as centralized model,
explained later), generators submit complex bids for each of their generating units. These bids are
supposed to reflect the cost characteristics of the unit (including the marginal, start-up and no-load
costs) as well as some technical parameters (minimum and maximum output, flexibility). Rather than
simply stacking the bids, the system operator then performs a central unit commitment that
determines the production schedule and the prices for an entire day divided in periods of half an hour
or an hour. For example, suppose a thermal generator with low marginal cost and high start up cost is
shut down temporarily. In the centralized dispatch system, the generator submits a complex bid
consisting of all details mentioned above. While doing a central unit commitment, the system operator
will not only consider its marginal cost, but will also take into account its start-up cost. On the other
hand, if the market is not centrally dispatched, only the marginal cost of the generator will be taken
into account to decide its selection or exclusion.

The advantage of complex bids is that they allow the system operator to take account, the true
characteristics of the generators and thus, potentially, do a more efficient job of minimizing the cost.
Setting the price becomes a disadvantage and requires a complex optimization problem to be
solved.This leads to higher cost of computation & lower transparency.

Day-Ahead Market and Real Time Market

As mentioned earlier, imbalances arise due to deviation between the forward contracted amounts and
the actual or near real time estimation of consumption. The spot markets are meant to provide a
mechanism for handling these imbalances. However, the energy trading must stop at some point
before real time to give the system operator to balance the system. How much time should elapse
between this gate closure and real time is a hotly debated issue.

Large coal based thermal plants take more than an hour to start up. Such generators can not bid if the
gate closure for spot market is lesser than one hour. Under these circumstances, it is beneficial to
have two energy markets: a Day Ahead (DA) market in addition to Real Time (RT) market. A DA
market, as its name implies, operates a day in advance of the RT market. A day-ahead market
becomes beneficial as follows:

First, it can be beneficial if generators have high start up costs and start and stop each day. In a
centralized dispatch model (explained later), the system operator integrates the start up costs of
these generators so as to come out with the start/stop decision in a longer term dispatch process. In
other words, the time horizon for optimizing dispatch decisions is a day, not an hour or less.

Second, it can be beneficial if generators would otherwise be able to game the market to lift spot
prices by withdrawing capacity at short notice - a form of market power. This form of market power is
explained with the help of Figure 3.12. Figure 3.12(A) shows competitive MCP denoted as MCP1. Now
suppose, if this generator company forms a coalition with Gen 7 and temporarily closes Gen 4 on the
terms of sharing the profit with Gen 7. Now, Gen 7 becomes the marginal generator and sets the MCP
(i.e., MCP2), which obviously is higher than the competitive MCP, as shown in Figure 3.12(B). If the
system operator needs to plan operations a few hours ahead of time and relies on generator promises
of availability, then under such cases, withdrawing capacity of a generator just ahead of real time
leads to calling of an expensive generation. The day-ahead contracts can remove the gaming incentive
from generators because their prices are locked in day ahead, and they can't play the same game in
the day ahead market because more alternatives are available day ahead than in real time.

The RT market provides volatile prices, in general. The DA market can promote demand response. If
the DA price is high, the loads can choose not to buy, and they have a day to plan for alternative
arrangements.

Figure 3.12: Change in MCP when capacity is withdrawn

MODELS FOR TRADING ARRANGEMENTS

As mentioned earlier, the decision about the market architecture is dominated by the factor - how
strongly the allied markets are integrated. The electrical energy takes different forms in different
markets. In other words, same electrical energy is valued as different products in different markets. If
the peculiarities associated with electricity would not have been there, then the four pillars of market
design would have been absent and electricity as a commodity would have been treated at par with
the other commodities. Then, the players in the market would only bid or play for price and quantity.
However, due to peculiarities of electricity, apart from energy market (bilateral or spot), other
markets also come into picture in which electricity, as a product, takes different forms. This fact
strongly influences the decision about the dispatch philosophies or in other words, the short-term
trading arrangements.

The dispatch philosophies are based on the degree to which the operational and commercial
arrangements for scheduling, imbalances, congestion and ancillary services are integrated with spot
markets. Depending upon how various markets associated with the four pillars are arranged, the
market dispatch procedures take the form of cascaded markets or integrated markets. Broadly
speaking, the integrated markets lead to economic efficiency at the cost of loss of transparency. On
the other hand, cascaded markets though inefficient, provide transparency. Keeping this in mind, the
dispatch philosophies or rather, the short-term trading arrangements can be classified into two broad
categories:
1. Integrated or Centralized Dispatch
2. Decentralized Dispatch

As the name suggests, the integrated model is integrated with strong linkages between various
aspects stated above. On the other hand, decentralized markets provide scattered efforts for various
arrangements in a power market.

One of the essential differences between integrated and decentralized markets is whether or not the
system operator administers a spot market integrated with the pricing of energy imbalances,
congestion management and ancillary services. The integrated model mandates the SO to run the spot
market, integrated with imbalances, and the others. On the other hand, the decentralized model
attempts to keep the spot market separate from the system operator, to be organized off-line by the
traders.

Integrated or centralized markets are now being commonly employed in USA. In this, the system
operator schedules forward contracts at the request of traders, but also takes bids from traders to
modify scheduled contracts and to provide energy imbalances, congestion management and ancillary
services. The system operator runs the spot market using large computer optimization program, and
by doing so, the system operator minimizes the overall cost of these services.

The decentralized model was employed in earlier Californian market (now it has moved towards
integrated model) and also in UK after adopting NETA3. In this model also, the system operator
schedules traders' contracts. However, the spot market is held separately and the decisions of the
same are conveyed to the system operator. The system operator has to administer arrangements for
imbalances. As far as possible, the traders run the spot market and manage congestion, while
separate arrangements are set up for ancillary services. The decentralized model requires not only
private markets for regular energy to cater to imbalances of forward markets, but also markets for
congestion energy and markets for ancillary services. As mentioned in [1], in a liquid and efficient
market, all these separate products will be exchanged at the same price, time and place. However, the
decentralized model does not ensure that the prices of all these different products converge. This may
be looked upon as an inefficiency. The integrated model, on the other hand, integrates energy of
imbalances, congestion, reserves and spot sales together and sells at the spot price determined by the
system operator, achieving economical efficiency in the dispatch process.

In the following sections, we intend to provide more details on the concepts of centralized and
decentralized markets, particularly comparing them on the following aspects:

1. Imbalance energy
2. Congestion Management
 3. Ancillary Services

Integrated or Centralized Model

The schematic of centralized dispatch model is presented in Figure 3.13. As shown, the joint
optimization of all markets is done by the system operator. It should be noted that the energy market
refers to the short term competitive market or the spot market, as defined in this module. As shown in
the figure, the buyers and sellers provide their bids and offers to the system operator. The buyers in
this model supply the complex bids. The system operator then performs the central unit commitment,
taking into account the complex bids. The system operator accomplishes this by solving a complex
optimization problem, typically known as Security Constrained Economic Dispatch, in USA market
context. The outcome of this dispatch is the nodal prices, popularly known as Locational Marginal
Prices (LMPs). If the losses are neglected and the network constraints are non-binding, the outcome of
this dispatch and that shown in Figure 3.11 carry the same meaning. In other words, all LMPs would
come out to be the same which means nothing but common MCP to all. However, this market clearing
is influenced and altered if the network capacities are congested and then the nodal LMPs would come
out to be different. In other words, the network congestion is implicit to this type of market
clearing.Some of the important features of this dispatch philosophy are discussed next.
 Figure 3.13: Centralized Dispatch

Treatment of Imbalances

How contracts and imbalances are tackled in the integrated model? If the system operator does the
central unit commitment using a complex optimization process, how are the forward contracts
accommodated? These are a couple of questions which need further explanation.

In the integrated dispatch model, all differences between contract positions and actual production,
consumption, regardless of cause, are traded at the market prices (spot prices) that come out of the
system operator's central optimization process and the forward contracts in the integrated model
remain financial in nature only. Whenever the traders make bilateral contracts in the integrated
structure, it is not necessary for the system operator to know anything about the bilateral or forward
contracts. This is because the system operator runs the least cost dispatch optimization program to
come out with locational spot prices and the settlements are done based on these locational spot
prices (LMPs). The effect of central optimization for least cost dispatch is that every generator is
scheduled by the system operator irrespective of its forward contract obligation. Thus, it may so
happen that a generator has contractual obligation of 20 MW and the system operator in fact, may
schedule this generator to produce zero MW! The effect of getting into a forward contract is then left
only as a risk hedging tool by locking in to some earlier decided prices. We will see more of this in the
module on risk hedging.

Sometimes due to operational constraints, the generation units are required to be scheduled, rather
than shutting them down. In this case, the generators are said to do self scheduling. In other words,
the system operator, while running its least cost optimization program, must schedule the MWs
offered by this generator. This is also known as inflexible bidding. The system operator, while running
its optimization program, shows this generator as a zero priced bid, so that it gets selected. Similarly,
for load, it shows the self scheduled load as an infinitely priced offer.

What is the effect of offering either a flexible or inflexible bid on revenues to generators? Let us see
the generator's perspective in case of flexible and inflexible bid submission in the spot market.
Suppose, a generator has a bilateral contract for 100 MW and its marginal price is INR 3000/MW.
Now, there are two choices for this generator.

1. Submit an inflexible bid. It can specify that, regardless of price, the system operator should
schedule this generator to inject 100 MW.
2. Submit a flexible bid. It can specify that anything up to its maximum capacity can be
dispatched by the system operator, as long as spot price exceeds its marginal price, i.e., INR
3000/ MW.

The first option essentially replicates the decentralized model. In this, the operator is meant to
schedule the bilateral transactions physically. However, in case 2, if the spot price falls below INR
3000/MW, the generator will not be dispatched by the system operator, as per the least cost dispatch
criteria. However, the requirement of load involved in the bilateral contract will still be satisfied. It is
equivalent to meeting the generator's contract by purchasing electricity (i.e., imbalances) in the spot
market at a lower rate than its running cost. Alternatively, if the spot price rises above INR 3000/MW,
the system operator will dispatch all MWs of this generator. It is easy to conclude that in either of the
circumstances, the generator is better off being flexible than being inflexible. Indeed, in practice, in
the integrated markets that are operating, much of the generation is offered as flexibly as its
production characteristics allow.

When all market participants are flexible, willing to modify operations from their contracted levels if
profitable, the system operator's dispatch is fully separate from forward contracts. The forward
contracts then become financial in nature only. The system operator will only know about a contract if
the traders involved have chosen to be inflexible. In case of all generators opting for flexible bids, it is
not relevant also for the system operator to know about the contract schedules.

Congestion Management

The process of congestion management is implicit to market operation of the integrated system.
Congestion is solved as an integral part of the calculation of the least cost dispatch, where cost is
defined by generator bids. The system operator uses the information regarding the bids and network
condition to determine the most economical way to operate the system within the physical constraints
using optimization software.

Pricing for congestion (i.e., the price charged for transporting electricity over scarce transmission) is
also straightforward. Traders who schedule contracts across valuable transmission lines are charged
for transmission usage which is equal to the energy price difference between the two ends of a
transaction. As mentioned earlier, the energy price at each node is calculated by the system operator
using central optimization process. Congestion management and its pricing are thus integrated with
energy pricing in the integrated model.

Ancillary Services
Though as good as 40 ancillary services can be listed, when it comes to classification of market
models based on ancillary services procurement, they essentially refer to the capacity of generators to
provide reserve. The reserves are not a separate service from energy, they are options to buy energy
if required. They should be priced as options to call energy in the spot market. However, complicating
factor is that the same generating unit can provide energy in the spot market, as well as can act as a
reserve. Hence, the system operator's dilemma is about how much of it should be scheduled in the
spot market and how much should be kept idle as a reserve.

In the integrated or centralized markets, depending on various technical criteria, the system operator
does a joint optimization of energy and reserve market so that optimum scheduling is done with
minimum cost as well as appropriate amount is kept for reserve in the optimal fashion. More details on
this issue are provided in the module on ancillary services management.

Decentralized Model

The decentralized and integrated models are most clearly distinguished by the different roles of
forward/bilateral contracts in the procedures used to schedule and dispatch generation. While the
integrated model treats the contracts essentially as financial agreements, and dispatches generators
to minimize overall costs, the decentralized model requires the system operator to schedule the
system explicitly using the contracts. Thus, transaction is treated as a basic unit to be accommodated
in real time system operations.

In all trading models, market participants can make and trade contracts in diverse markets separate
from the system operator. The contracts could be one-to-one contracts or obtained through an
organized trade. At some predetermined moment prior to real time operations, however, the system
operator has to take over to deliver the contracts. The system operator is not intended to facilitate a
spot market - he simply schedules trades that have been arranged elsewhere. While transferring
transactions to the system operator for scheduling, the condition that the amount bought should be
equal to sold should be satisfied. Each seller must have a buyer and each buyer a seller. The aim of
decentralized model is to leave as much of the trading as possible to the traders, whereas, in the
integrated model, on the day ahead and in real time, the system operator makes the trades by
following instructions incorporated in the traders' bids.

The schematic of decentralized dispatch model is presented in Figure 3.14. As shown, the energy
market is not an integral part of system operator's activity and essentially depends on an external
activity exclusive from the system operator.
 Figure 3.14: De-centralized Dispatch

Treatment of Imbalances

The participants of various forward trades wish to balance their positions near real time. This is
generally accomplished through a spot market. This provides a common clearing price for imbalances,
which is competitive in nature. If a decentralized system has market based imbalance prices, then that
price becomes the price at which the system operator will buy or sell energy. The market based price
of imbalances provides a reference price for forward contracts. When participants sign contracts, the
contract prices are directly compared with the expected imbalances prices as imbalance provides
direct substitute for contract energy. The spot market for the imbalance energy is generally run by a
power exchange, where the participants submit simple price-quantity bids rather than complex bids.

Congestion Management
Congestion management in decentralized model can occur in one of the following ways:

1. Allocation of transmission rights on pro-rata basis

2. Allocation of transmission rights on first come first serve basis
3. Auction of available transmission capacity
4. Special case of zonal pricing with market splitting
5. Pro-rata curtailment in case of contingent situation

The first two approaches are not essentially the market based solutions. They don't reflect the
willingness of a trader to pay for obtaining transmission rights. These methods do not take into
account the network element interdependencies. On the other hand, the third option, i.e., of
auctioning of capacity rights reflects traders' willingness to pay. The fourth option, i.e., special case of
zonal pricing, separates markets across the transmission bottleneck and imposes a congestion fee for
a bilateral transaction taking place across the transmission bottleneck. The last approach, i.e.,
curtailment, is more of congestion alleviation technique invoked in real time.

For all these methods, a different solution needs to be worked out to manage the congestion in real
time. In the decentralized model, the system operator will only deviate from the final contract
schedules in the dispatch if he needs to do so in order to maintain security, and may not make
efficient trades even if the traders ask him to since there is no bidding mechanism for them to do so.
It should be noted that congestion management is not implicit with market clearing process.
Transmission capacity allocation needs to be done explicitly.

Ancillary Services

Unlike in the centralized model, the system operator procures all types of ancillary services generally
by making long term bilateral contracts with generator. It is obvious that this is not the best possible
way of obtaining reserves, though it gets rid of complexities associated with joint optimization of
energy and reserves market and lack of transparency associated with it. There is a possibility that the
reserves are procured on market basis. However, in decentralized system, it takes the form of
cascaded markets. In other words, after passing through forward contracts and spot market, the
generators provide rest of their capacity to reserves markets, subject to technical compatibility, with
the needs. This arrangement can take the other form in the sense, the generators get involved in long
term contracts for reserves and after subtracting for forward energy contracts, the rest is offered in
the spot market for balancing. There is no joint optimization as is done in the centralized dispatch
market.

Comparison at a glance

The conclusion of the discussion on centralized dispatch and decentralized dispatch is provided in
Table 3.3 by comparing the various aspects of two models.

ISO MODEL OR TSO MODEL

In section 3.2, we have seen how various trading arrangements take different forms depending on the
interaction between various entities of the market. In all the four models presented in section 3.2, the
issue of ownership of transmission and distribution network is not discussed. The restructured power
system models across the world can also be classified according to the ownership of transmission
network. Rather, more clear distinction would be based on whether the system operator itself is owner
of the transmission network or somebody else owns it. In ISO (Independent System Operator) model,
the owner of transmission network is different from the system operator. In TSO (Transmission
System Owner) model, system operator itself owns the infrastructural investments. This arrangement
is seen in most of the developing countries.

ISO Model
ISO model is practiced in those countries in which transmission companies are also providing the
generation and distribution services in their area of operation. Further, in these countries, sufficient
number of equal sized transmission companies exists in the market and it is not possible to club the
system operation function with any of these companies for commercial reasons. Therefore, separation
of ownership of the transmission assets from the system operation function is considered necessary to
avoid any preferential treatment for dispatching its own generation.

No Atrribute Centralized Model Dcentralized Model

1 Unit Commitment Central Individual

Reservw market Separate reserve market

2 Reserve integrated with spot or obtained through long
market term contracts

Individual schedules
Basis for Bids abd offers of
3 arising out of bilateral
scheduling participants
transactions

Integrated with spot Generally through day-

4 Imbalances
market ahead market

Involvement of
5 system operator in Yes No
day- ahead market

Congestion
6 Implicit Explicit
Management

Significance of
7 Risk hedging Physical obligation
forward contracts

SMD market in USA NETA in UK Nordic

8 Example
like PJM, ISO-NE pool

Table 3.3: Comparison between centralized and decentralized dispatch models

TSO Model

In TSO model, operation and ownership of the grid are integrated into a single entity which is
responsible for development of transmission system and to provide non-discriminatory open access to
all eligible market participants. It is also responsible for system operation functions. Neutrality is an
important aspect of the TSO to ensure an efficient market. This model is prevalent in the whole of the
Europe.
 Module 4: Transmission Congestion Management
INTRODUCTION

Congestion management in a multi-buyer/ multi-seller system is one of the most involved tasks if it
has to have a market based solution with economic efficiency. In a vertically integrated utility
structure, activities such as generation, transmission and distribution are within direct control of a
central agency or a single utility. Generation is dispatched in order to achieve the system least cost
operation. Along with this, the optimal dispatch solution using security constrained economic dispatch
eliminates the possible occurrence of congestion. This effectively means that generations are
dispatched such that the power flow limits on the transmission lines are not exceeded.

One should not expect things to be as simple in a deregulated power environment. In a deregulated
environment, every buyer wants to buy power from the cheapest generator available, irrespective of
relative geographical location of buyer and seller. As a consequence of the this, the transmission
corridors evacuating the power of cheaper generators would get overloaded if all such transactions are
approved. Congestion is then said to have occurred when system operator finds that all the
transactions can not be allowed on account of overload on the transmission network. Congestion
management is a mechanism to prioritize the transactions and commit to such a schedule which would
not overload the network. Despite these measures, congestion can still occur in real time following a
forced outage of transmission line. The system operator then handles this situation by means of real
time congestion management. Thus, congestion management involves precautionary as well as
remedial action on system operators part, as follows:

Allow only that set of transactions which, taken together, keeps the transmission system
within limits.
Even if this care is taken, in real time, the transmission corridors may get overloaded due to
unscheduled flows. The system operator has to take some remedial action.

The scope of transmission congestion management in the deregulated environment involves defining a
set of rules to ensure control over generators and loads in order to maintain acceptable level of
system security and reliability. The rules should ensure market efficiency maximization with short
term as well as long term horizons. The robustness of rule set is important as under open market
structure a set of players will always be looking for loopholes in the mechanism to exploit it.

In a deregulated structure, the market must be modeled so that the market participants (buyers and
sellers of energy) engage freely in transactions and play as per market forces, but in a manner that
does not threaten the security of the power system. Thus, irrespective of the market structure in
place, congestion management has universally become an important activity of power system
operators. Universally, the dual objectives of congestion management schemes have been to minimize
the interference of the transmission network in the market for electrical energy and to simultaneously
ensure secure operation of the power system.

This chapter is aimed at explaining various congestion management schemes employed under various
market structures. The chapter provides classification of congestion management schemes based on
the economic efficiency associated with each one of them.

Definition of Congestion

Whenever the physical or operational constraints in a transmission network become active, the system
is said to be in a state of congestion. The possible limits that may be hit in case of congestion are: line
thermal limits, transformer emergency ratings, bus voltage limits, transient or oscillatory stability, etc.
These limits constrain the amount of electric power that can be transmitted between two locations
through a transmission network. Flows should not be allowed to increase to levels where a
contingency would cause the network to collapse because of voltage instability, etc.

The peculiar characteristics associated with electrical power prevent its direct comparison with other
marketable commodities. First, electrical energy can not be stored in large chunks. In other words, the
demand of electric power has to be satisfied on a real time basis. Due to other peculiarities, the
flexibility of directly routing this commodity through a desired path is very limited. The flow of electric
current obeys laws of physics rather than the wish of traders or operators. Thus, the system operator
has to decide upon such a pattern of injections and take-offs, that no constraint is violated.

How Transfer capability is limited?

Congestion, as used in deregulation parlance, generally refers to a transmission line hitting its limit.
The ability of interconnected transmission networks to reliably transfer electric power may be limited
by the physical and electrical characteristics of the systems including any or more of the following:

Thermal Limits: Thermal limits establish the maximum amount of electrical current that a
transmission line or electrical facility can conduct over a specified time period before it
sustains permanent damage by overheating.
Voltage Limits: System voltages and changes in voltages must be maintained within the
range of acceptable minimum and maximum limits. The lower voltage limits determine the
maximum amount of electric power that can be transferred.
Stability Limits: The transmission network must be capable of surviving disturbances
through the transient and dynamic time periods (from milliseconds to several minutes,
respectively). Immediately following a system disturbance, generators begin to oscillate
relative to each other, causing fluctuations in system frequency, line loadings, and system
voltages. For the system to be stable, the oscillations must diminish as the electric system
attains a new stable operating point. The line loadings prior to the disturbance should be at
such a level that its tripping does not cause system-wide dynamic instability.

The limiting condition on some portions of the transmission network can shift among thermal, voltage,
and stability limits as the network operating conditions change over time. For example, for a short
line, the line loading limit is dominated by its thermal limit. On the other hand, for a long line, stability
limit is the main concern. Such differing criteria further lead to complexities while determining transfer
capability limits.

Importance of congestion management in the deregulated environment

If the network power carrying capacity is infinite and if there are ample resources to keep the system
variables within limits, the most efficient generation dispatch will correspond to the least cost
operation. Kirchoffs laws combined with the magnitude and location of the generations and loads, the
line impedances and the network topology determine the flows in each line. In real life, however, the
power carrying capacity of a line is limited by various limits as explained earlier. These power system
security constraints may therefore necessitate a change in the generator schedules away from the
most efficient dispatch. In the traditional vertically integrated utility environment, the generation
patterns are fairly stable. From a short term perspective, the system operator may have to deviate
from the efficient dispatch in order to keep line flows within limits. However, the financial implications
of such re-dispatch does not surface because the monopolist can easily socialize these costs amongst
the various participants, which in turn, are under his direct control. From planning perspective also, a
definite approach can be adopted for network augmentation.

However, in deregulated structures, with generating companies competing in an open transmission

access environment, the generation / flow patterns can change drastically over small time periods with
the market forces. In such situations, it becomes necessary to have a congestion management
scheme in place to ensure that the system stays secure. However, being a competitive environment,
the re-dispatch will have direct financial implications affecting most of the market players, creating a
set of winners and losers. Moreover, the congestion bottlenecks would encourage some strategic
players to exploit the situation. The effects that congestion is likely to cause are discussed next.

Effects of Congestion

The network congestion essentially leads to out-of-merit dispatch. The main results of these can be
stated as follows:

Market Inefficiency: Market efficiency, in the short term, refers to a market outcome that
maximizes the sum of the producer surplus and consumer surplus, which is generally known
as social welfare. With respect to generation, market efficiency will result when the most cost-
effective generation resources are used to serve the load. The difference in social welfare
between a perfect market and a real market is a measure of the efficiency of the real market.
The effect of transmission congestion is to create market inefficiency.
Market Power: If the generator can successfully increase its profits by strategic bidding or by
any means other than lowering its costs, it is said to have market power. Imagine a two area
system with cheaper generation in area 1 and relatively costlier generation in area 2. Buyers
in both the areas would prefer the generation in area 1 and eventually the tie-lines between
the two areas would start operating at full capacity such that no further power transfer from
area 1 to 2 is possible. The sellers in area 2 are then said to possess market power. By
exercising market power, these sellers can charge higher price to buyers if the loads are
inelastic. Thus, congestion may lead to market power which ultimately results in market
inefficiency.

In multi-seller / multi-buyer environment, the operator has to look after some additional issues which
crop up due to congestion. For example, in a centralized dispatch structure, the system operator
changes schedules of generators by raising generation of some while decreasing that of others. The
operator compensates the parties who were asked to generate more by paying them for their
additional power production and giving lost opportunity payments to parties who were ordered to step
down. The operator has to share additional workload of commercial settlements arising due to network
constraints which, otherwise, would have been absent.

One important thing to be noted is that creation of market inefficiency arising due to congestion in a
perfectly competitive market acts as an economic signal for network reinforcement. The market design
should be such that the players are made to take a clue from these signals so as to reinforce the
network, thus mitigating market inefficiency.

Desired Features of Congestion Management Schemes

Tackling the congestion problem takes different forms in different countries. It really depends on what
type of deregulation model is being employed in a particular region. Certain network topologies,
demographic factors and political ideologies influence the implementation of congestion management
schemes in conjunction with overall market design.

Any congestion management scheme should try to accommodate the following features:

Economic Efficiency: Congestion management should minimize its intervention into a

competitive market. In other words, it should achieve system security, forgoing as little social
welfare as possible. The scheme should lead to both, short term and long term efficiency. The
short term efficiency is associated with generator dispatch, while long term efficiency pertains
to investments in new transmission and generation facilities
Non discriminative: Each market participant should be treated equally. For this, the network
operator should be independent of market parties and he should not derive any kind of benefit
 from occurrence of congestion. Otherwise it provides perverse signals for network expansion.
Be transparent: The implementation should be well defined and transparent for all
participants.
Be robust: Congestion management scheme should be robust with respect to strategic
manipulation by the market entities. This again refers back to principle of economic efficiency

Though a variety of forms of congestion management schemes are practiced throughout the power
markets of the world, the nodal pricing or the optimal power flow based congestion management
scheme is said to satisfy most of the desired features of the same, especially the feature of economic
efficiency. Each practiced method has strengths and flaws and also interrelationships to some extent.
Each maintains power system security but differs in its impact on the economics of the energy market.

CLASSIFICATION OF CONGESTION MANAGEMENT

MECHANISMS

The congestion management schemes are strongly coupled with the overall market design. Efficient
allocation of scarce transmission capacity to the desired participants of the market is one of the main
objectives of congestion management schemes. Thus, distinction among them can be made based on
market based congestion management methods and other methods. Market-based solutions to
congestion are deemed fairer as they contribute better to economic efficiency than other methods.
Methods other than market based make use of some criteria to allocate the transmission capacity.
These methods are supposed to introduce some kind of arbitrariness as they do not contribute
towards efficient pricing of congested link. Classification of congestion management schemes on these
lines is shown in Table 4.1.

Non - market Methods Market Based Methods

Explicit Auctioning of
1 Type of contract 1
network capacity

Nodal pricing (OPF

First come first
2 2 based congestion
serve
management)

3 Pro - rata methods 3 Zonal pricing

Price area congestion

4 Curtailment 4
management

5 Re - dispatch

6 Counter trace

Table 4.1: Classification of congestion management schemes

Before moving on to see details of these methods, we first see various phases of the entire process of
congestion management on timeline. This is explained with the help of Figure 4.1. As the transactions
keep on committing, the system operator continuously updates the available transfer capability
between various regions / areas / nodes in the system. This becomes essential because as the day-
ahead (or the spot) market approaches, the operator should have knowledge about the network
capacity left for settling the market. The transmission network capacity allocation in a coordinated
market may take an explicit or implicit form. In other words, there can be a separate market for
transmission capacity reservation or it may be integrated with the coordinated market.

Even after capacity allocation, the real time flows may lead to violation of transmission capacities. In
order to relieve congestion during real time, congestion alleviation methods are employed.

Figure 4.1: Phases of network access with respect to congestion

Out of several congestion management techniques listed above, following are exclusively termed as
congestion alleviation methods:

1. Re-dispatch
2. Counter Trade
3. Curtailment

It should be noted that the capacity allocation methods usually allocate the transmission capacity in
ex-ante manner before physical delivery of energy. On the other hand, congestion alleviation methods
are termed as remedial actions. The procedure of capacity allocation starts with the calculation of
Available Transfer Capability (ATC). Let us see some details of ATC calculation in the next section

CALCULATION OF AVAILABLE TRANSFER CAPABILITY (ATC)

In a vertically integrated market, the inter-area tie lines are designed only to address the reliability,
system security and system restoration purposes. This integration of various systems becomes a
market need in the deregulated era. Thus, inter-area tie lines become means of bulk power transfers
on a regular basis from sources of cheap generation to loads. In other words, due to deregulation, the
paradigm of grid integration has shifted from regional self-sufficiency to optimal utilization of
resources across large geographical areas. Thus, it becomes imperative on the part of system operator
to quantify the Available Transfer Capability (ATC) of the network and allocate the same to the market
participants in an efficient manner.

Generally, the non-market based methods rely upon the information about the ATC in order to take a
decision while allowing the next set of transactions. Thus, calculation of ATC gains a lot of importance
under such market structures. In the early days of deregulation in USA, the ATC values for the next
hour and for each hour into the future would be placed on a website known as the open access same-
time information system (OASIS), to be operated by the ISO. Anyone wishing to send a power
transaction on the ISO's transmission system would access OASIS web pages and use the ATC
information available there to determine if the transmission system could accommodate the
transaction, and to reserve the necessary transmission service.

Next section provides formal definitions of some of the common terms pertaining to ATC.

Definition of Various Terms

Available Transfer Capability (ATC

It is a measure of the transfer capability remaining in the physical transmission network for further
commercial activity over and above already committed uses. Mathematically, ATC is defined as the
Total Transfer Capability (TTC) less the Transmission Reliability Margin (TRM), less the sum of existing
transmission commitments (which includes retail customer service) and the Capacity Benefit Margin
(CBM
ATC = TTC - TRM - Existing Transmission Commitments (including CBM)

Total Transfer Capability (TTC)

It is defined as the amount of electric power that can be transferred over the interconnected
transmission network in a reliable manner while meeting all of the specific set of defined pre and post
contingency system conditions.

Transmission Reliability Margin (TRM)

It is defined as the amount of transmission transfer capability necessary to ensure that the
interconnected transmission network is secure under a reasonable range of uncertainties in system
conditions.

Capacity Benefit Margin (CBM)

It is defined as the amount of transmission transfer capability reserved by load serving entities to
ensure that the interconnected systems do meet generation reliability requirements.

The NERC report [24] brings out the difference between transfer capability and transmission
capacity. According to this report, the capacity' specifically mentions the rating of the equipment, for
example, the ampacity of the conductor. On the other hand, the capability' depends upon generation,
customer demand and the conditions in a transmission system for the given time period. Thus, the
capacity' of a circuit may not change much from time to time. However, the capability' always
changes with the time by virtue of changes in the system condition.

As mentioned earlier, ability of the network system to reliably deliver power is limited by physical and
electrical characteristics of the system. These limits are: Thermal, Voltage and Stability. During the
varying conditions of power system, one of these limits plays a major role in deciding the transfer
capability. Determining which limit is binding during a particular time is a challenging task and makes
computation of ATC an involved task.

Many methods have been suggested to calculate the ATC. The methods differ on the basis of the
power flow model being employed, the system aspects considered, the compelling limits under
consideration and some other factors. However, a broad way of classifying methods is based on the
type of limit considered, i.e., Thermal limit, Voltage limit or the Angular stability limit. The DC power
flow methods take into consideration only the thermal limits. The AC Optimal power flow (OPF)
methods consider thermal as well as voltage limits. Then, there is another version called Continuation
power flow method (CPF). It considers a series of power system solutions to be solved and tested for
limits. The amount of transfer is gradually increased from the base case until a binding limit is
encountered. Ejebe et al. [31] have described a method based on continuation power flow,
incorporating limits of reactive power flows, voltage limits, as well as voltage collapse and line flow
limits. Stability constrained methods require transient studies to be carried over a case developed with
anticipated scenario. Christie et al. [1] have proposed a method based on DC power transfer
distribution factors (PTDF). This utilizes DC load flow based formulation, and computation of
simultaneous ATC has also been considered using an optimization based approach. Let us see the
details of DC power flow based ATC calculation next.

ATC Calculation using PTDF and LODF based on DC Model

One way of calculating ATC from node A to node B is to use DC load flow (explained later) repetitively
by increasing the amount of transaction until a limit of any of the corridor is reached. However, this is
computationally inefficient. Instead, the Power Transfer Distribution Factor (PTDF) can be used to
calculate the maximum allowable flow for a given pair of injection and take-off points. It is also
necessary to consider the effects of contingencies like line outages. This can be achieved using Line
Outage Distribution Factor (LODF). Let us first see the details of DC load flow model.

DC Load Flow Model

Following are the assumptions when DC model is employed instead of AC model:

Voltage magnitudes are constant.
Only angles of complex bus voltages vary.
The variation in angle is small.
Transmission lines are lossless.

These assumptions create a model that is a reasonable first approximation for the real power system,
which is only slightly nonlinear in normal steady state operation. The model has advantages for speed
of computation, and also has some useful properties like linearity and superposition.

With these assumptions, power flows over transmission lines connecting bus i and bus j is given as:

....................................................................................................................................(4.1)
Where,

line inductive reactance in per unit

phase angle at bus l

phase angle at bus m

The total power flowing into the bus i, Pi, is the algebraic sum of generation and load at the bus and is
called a bus power injection. Thus,

......................................................................................................................(4.2)
This can be expressed in a matrix form as:
 .................................................................................................................................(4.3)
Where, the elements of the susceptance matrix BX are functions of line reactances . One node is
assigned as a reference node by making its angle zero and deleting corresponding row and column in

matrix. Thus,

............................................................................................................................(4.4)

The dimension of obtained is . Let us augment it by adding zero column and row
corresponding to reference bus. The angles in equation 4.3 can be found out as

.................................................................................................................................(4.5)
Thus, power flow over line lm can be found out using equation 4.1.

Power Transfer Distribution Factor (PTDF)

From the power transfer point of view, a transaction is a specific amount of power that is injected into
the system at one bus by a generator and drawn at another bus by a load. The coefficient of linear
relationship between the amount of a transaction and flow on a line is represented by PTDF. It is also
called sensitivity because it relates the amount of one change - transaction amount - to another
change - line power flow.

PTDF is the fraction of amount of a transaction from one bus to another that flows over a transmission

line. is the fraction of a transaction from bus i to bus j that flows over a transmission line
connecting buses l and m.

.....................................................................................................................(4.6)

Calculation of PTDF Using DC Model

Suppose there exists only one transaction in the system. Let the transaction be of 1 MW from bus i to

bus j. Then, the corresponding entries in equation 4.7 will be: and . All other entries will
be zero. From equation 4.5, we get
 ..........................................................................................................................(4.7)
Similarly,

...........................................................................................................................(4.8)
Thus,

............................................................................................................................(4.9)

.......................................................................................................................(4.10)
Using equations 4.9, 4.10 and 4.1, the PTDF can be calculated as

...................................................................................................(4.11)

Reactance of transmission line connecting buses l and m

Entry lth row and ith column of the bus reactance matrix X
The change in line flow associated with a new transaction is then

...................................................................................................................(4.12)
Where,
l and m buses at the ends of the line being monitored
i and j from and to bus numbers for the proposed new transactions

New transaction MW amount

ATC calculation Using PTDF

ATC is determined by recognizing the new flow on the line from node l to node m, due to a transaction
from node i to node j. The new flow on the line is the sum of original flow and the change.

............................................................................................................(4.13)

Where, is the base case flow on the line and is the magnitude of proposed transfer. If the limit

on line lm, the maximum power that can be transferred without overloading line lm, is , then,

....................................................................................................................(4.14)

is the maximum allowable transaction from node i to node j constrained by the line from node l
to node m. ATC is the minimum of the maximum allowable transactions over all lines.

Using the above equation, any proposed transaction for a specific hour may be checked by calculating
ATC. If it is greater than the amount of the proposed transaction, the transaction is allowed. If not,
the transaction must be rejected or limited to the ATC.

.......................................................................................................(4.15)
Using the above equation, any proposed transaction for a specific hour may be checked by calculating
ATC. If it is greater than the amount of the proposed transaction, the transaction is allowed. If not,
the transaction must be rejected or limited to the ATC.

Numerical Example of ATC Calculation Using PTDF

Consider a sample 3 bus system as shown in Figure 4.2.

 Figure 4.2: Sample 3 bus system

Bus Load
Generation
No (MW)
1 200 0

2 700 100

3 0 800

Table 4.2: Bus data for sample system

Line Max. Power

From To
Bus Bus
Reactance capacity
(pu) (MW)
1 2 0.1 600

2 3 0.033 200

3 3 0.1 600

Table 4.3: Line data for sample system

If we treat bus 1 as the reference bus, then the matrix X is obtained as

 ......................................................................................................................(4.16)

.....................................................................................................................(4.17)

Now let us calculate PTDF for a transaction between bus 1 and 3. Thus, PTDF on various corridors
using equation 4.11 can be given as:

............................................................................................................(4.18)

...........................................................................................................(4.19)

.........................................................................................................(4.20)
Similar calculations are done for transaction between buses 2 and 3, the PTDFs are given as

........................................................................................................(4.21)

.........................................................................................................(4.22)

........................................................................................................(4.23)
Now suppose, there are two transactions, one of 200 MW between buses 1 and 3, and the other of
600 MW between bus 2 and 3. Then, the power flow on all corridors due to these transactions can be
given as:
 ..............................................................................................(4.24)

The above equation establishes the base case flows of a system with two transactions in place. Now
let us calculate ATC for transferring power between bus 1 and 2 and also between bus 2 and 3. In
order to calculate ATC, we make use of equations 4.11, 4.12 and 4.13. First let us calculate ATC
between buses 1 and 3. Using equation 4.12,

.............................................................................................. .(4.25)

................................................................................................ .(4.26)

.................................................................................................(4.27)
Using equation 4.13, ATC between buses 1 and 3 is given as 179 MW. Similarly, for ATC between
buses 2 and 3,

...............................................................................................(4.28)

.................................................................................................(4.29)

...............................................................................................(4.30)
As per equation 4.13, ATC between buses 2 and 3 is 199.3 MW.

ATC Calculation Using PTDF and LODF

The calculation of ATC should also take into account the effect of line contingency. In other words, it
should indicate the available transfer capacity after considering the changes in line flows due to the
largest line contingency. For this, the concept of Line Outage Distribution Factor (LODF) is introduced
and used for ATC calculation.

Line Outage Distribution Factor (LODF):

When an outage occurs, the power flowing over the outaged line is redistributed onto the remaining

lines in the system. The LODF is the measure of this redistribution. is the fraction of the
power flowing on the line rs before it is outaged, which now flows over a line from l to m.

...........................................................................................................................(4.31)
The LODF is given by

..............................................................................................(4.32)
Where,

reactance of line connecting bus l and m

entry in lth row and rth column of bus reactance matrix X

number of circuits connecting bus l and bus m

Consider a transaction from bus i to bus j and the outage of a line from bus r to bus s (line rs). The
change in flow on line rs due to the transaction is

.......................................................................................................................(4.33)

When line rs is outaged, part of the flow appears on line lm. s resulting from both the outage of the
line rs and a new transaction from bus i to bus j is given by

..........................................................................................(4.34)

The maximum contingency limited transfer from bus i to bus j, limited by line lm, with the outage of
line rs, is given by

................................................................................................(4.35)

Where, indicates the post contingency flow limit on line lm.

To find the contingency limited ATC, all possible combinations of outaged lines and limiting lines must
be checked, as well as steady state transfer limit.

.............................................................................................(4.36)
Using the above equations, any proposed transaction for the specific hour may be checked by
calculating the ATC. If it is greater than the amount of proposed transaction, the transaction is
allowed. If not, the transaction must be rejected or limited to the ATC.

Calculation of PTDF Using AC Model

In the previous section we have seen PTDF calculation using DC power flow model. But this involves
many assumptions which lead to inaccurate results. More accurate PTDFs can be calculated using AC
power flow model. Line power flows are simply function of the voltages and angles at its terminal
buses. So PTDF is a function of these voltage and angle sensitivities.

Consider an n node system with nodes1,......,g as PV nodes (generator buses) and g+1,...,n as the PQ
nodes (load buses). Bus 1 is taken as slack bus. A transaction is defined by a set of four parameters
(t, i, j, Pt) where t is the transaction number, i and j are the source and sink nodes and Pt is the MWs
transacted. The change in flow of an arbitrary line lm can be evaluated by sensitivity analysis as
follows.

.....................................................................................(4.37)
From the converged base case Load Flow solution we have,

....................................................................................................................(4.38)
where J is load flow Jacobian. For a MW power transaction number t,

...............................................................................................................................(4.39)
 ..............................................................................................................................(4.40)

.................................................................................................................................(4.41)

................................................................................................................................(4.42)
where, . Substituting 4.6-4.9 in 4.5 and then in 4.4,

..................................................................................(4.43)

NON-MARKET METHODS OF CONGETSION MANAGEMENT

The non-market methods of congestion management essentially refer to network capacity allocation
based on some pre-defined set of rules that neglect the ability or the willingness of a player to pay for
the transmission capacity. For such schemes and also for explicit market based methods, the system
operator is required to know the capacity remaining with the grid, after accommodating the
transactions which is nothing but ATC.

Capacity Allocation on First Come First Serve Basis

There are some systems in which the bilateral contracts are awarded for transmission network access
on first come first served basis. The calculation of ATC facilitates a participant to determine whether
there is enough capacity available for him to do the transaction between two nodes of concern. If
enough capacity is left on the network so as to make a transaction, the participant books his
transaction with the system operator. After this, the system operator updates the ATC. The next
transaction in line again checks whether there is enough corridor capacity available to do the
transaction. The drawback associated with this mechanism is that the willingness to pay for
transmission usage is not taken into account. Those participants with high valuation of transmission
network may not get scheduled.

Capacity Allocation based on Pro-rata Methods

Various norms can be set to assign network capacities on pro-rata basis. The capacities can be
allocated on average load or generation, or percentage of long term transactions or maximum
demand, etc. In other words, all participants receive an equal percentage of the total amount of
capacity they apply for. These norms are used for capacity allocation as well as for congestion
alleviation, which is used in real time. This scheme also has same limitations as in first come first
served method.

Another limitation of this method is possible strategic behavior of the market participants. The system
of pro-rata distribution of capacity can lead to market parties applying for transmission capacity much
more than what they want, knowing that the actual amount that they will receive is physically limited.

Capacity Allocation based on Type of Contract

In this type of capacity allocation, network capacity is allocated to a particular type of transactions.
For example, capacity is first allocated to firm or long term transaction and in the rest of capacity,
maximum possible number of short term transactions are accommodated. If in the real time
operation, the re-dispatching of injections is required to be done, the short term transactions are
curtailed first ahead of long term or firm transactions.

EXPLICIT AUCTIONING

The principle of explicit auctioning is based on selling the available capacity of the tie line to the
highest bidder through auction. This is nothing but auctioning of the tie line capacity. The explicit
auctioning separates the energy market from transmission capacity market.

This approach is commonly used in Europe for capacity allocation at several borders. In explicit
auctioning, the system operators (or the TSO in Europe) determine ex-ante, the available transmission
capacity (ATC) considering security analysis, accepts bids from potential buyers and allocates the
capacity to the ones that value it most. Thus, explicit auctioning is a market based concept, which
provides economic signals. Thus, with perfect foresight, bidders for transmission capacity would
predict the electricity market outcome with efficient use of transmission.

A limitation of this mechanism is the increased complexity which may complicate trading activities of
market participants. Another limitation is that the mechanism fails to account for parallel flows in
meshed networks. In this context, a new method has been proposed called coordinated auctioning.

Coordinated Auctioning

The coordinated auctioning splits the markets into energy market and transmission capacity market.
Participants have to ensure that they own sufficient transmission rights to conclude their energy
exchanges. However, coordinated auctioning tries to overcome problems associated with explicit
auctioning by accounting for the effects of loop flows in the network. A central auctioneer is introduced
who manages capacity allocation at all borders included in the Internal European Market (IEM). For
coordinated auctioning, three steps are necessary:
Each system operator informs the central auctioneer about the available transmission
capability.
Market participants submit their bids to the central auctioneer.
The auctioneer allocates transmission capacity using a model similar to nodal pricing.

Market participants may value their willingness to pay for transmission rights by comparing the
different zonal prices. Hence, rational bidders for transmission rights will submit bids equal to the
zonal difference in energy prices, as savings for cheaper energy are traded off against the
transmission costs. With perfect foresight and all information available, a coordinated auction will lead
to the same allocation as the nodal pricing approach, which in turn is considered as an economically
efficient decision.
 NODAL PRICING: OPF BASED CONGESTION MANAGEMENT

The general idea of nodal pricing is to model an electricity market with its various economical and
technical specifications, such as generators' cost functions, demand elasticity, generation limits, line
power flow limits and optimize the system for maximizing social welfare. This problem represents one
of the commonly employed formulations of Optimal Power Flow (OPF). The name OPF does not stand
for any specific optimization problem, rather a number of optimization problems falls into the OPF
category. The basic aim of an OPF analysis is to reach an optimum power transfer situation without
violating the network constraints. In other words, the congestion management problem is tackled by
solving an optimization problem, with a set of constraints representing network constraints.

One of the outcomes of this optimization problem is the price at each node known as nodal price. It
reflects the temporal and spatial variation of energy price relating to the demand on that node.Nodal
pricing can be interpreted as a fully coordinated implicit auction. The market participants do not
explicitly participate into auctions for transmission capacity. On the contrary, they submit bids for
energy injection and take-off and they are scheduled such that transmission capacity is implicitly
allocated to them based on their energy bids, without violating the network constraints.

The practical OPF uses a formulation wherein ac power flow equations are added to the economic
dispatch as equality constraints with inequality constraints involving the flow MW, MVA or current on a
transmission line and voltages at a substation bus. A version of OPF is developed that takes into
account various contingencies referred to as a security constrained OPF or SCOPF.

OPF problems are formulated with a number of objectives. A brief list is outlined here:
Minimize the total cost of production
Maximize total social welfare
Minimize total system loss
Minimize the re-dispatch cost
Minimize the total adjustment
Minimize load curtailment

In the text to follow, we will discuss the formulation of AC OPF as well as DC approximation of it - DC
OPF.

DC OPF

This section introduces the formulation of DCOPF. Conventional Economic Dispatch problem can be
represented as a minimization of total generation cost as follows:

................................................................................................(4.44)

...............................................................................................(4.45)

................................................................................................(4.46)
 Where,

power output of generator i

generator i lower limit

Total system load

cost function of generator i

This problem can be converted into an OPF problem with inclusion of following:

The constraints representing the DC power flow equations

A set of inequality constraints representing line MW limits

The DC power equations as constraints are obtained by rearranging equation 4.3. This is given as:

..............................................................................................(4.47)
The line flow constraints are given as:

...................................................................................................(4.48)
This can be expressed in terms of phase angles. With addition of a slack variable, it can be given
as:

................................................................................................(4.49)

The slack variable has a lower limit of zero and an upper limit of . Thus, the simplified DC OPF
formulation can be given by following set of equations:

.....................................................................................................(4.50)
Subject to

..................................................................................................(4.51)
 ................................................................................................(4.52)

.................................................................................................(4.53)

The variables in this OPF are: bus phase angles , generator power outputs , m line flow limit
slack variables sij , where m is number of lines.

Inclusion of Load Elasticity in OPF

Elasticity of a load is represented by a benefit function . It represents the prices that the

load is willing to pay to purchase an amount of power and represents load's bid to the pool or
ISO. The objective function of the OPF problem will be modified as follows:

.............................................................................................(4.54)
The DC power flow equations of 4.31 are also modified to include the price elastic load as follows:

............................................................................................(4.55)
This formulation gives rise to additional set of variables, i.e., . Rest of the problem remains the same.

AC OPF

AC OPF gives a more realistic solution as losses are implicitly involved in the formulation itself. The AC
OPF formulation is given as follows:

..............................................................................................(4.56)
subject to, load balance equations as equality constraints

...........................................................(4.57)

........................................................(4.58)

And the inequality constraints

....................................................................................(4.59)

....................................................................................(4.60)

....................................................................................(4.61)

....................................................................................(4.62)

.....................................................................................(4.63)

......................................................................................(4.64)

Where,

Vi......voltage at ith bus

......real power flow on line km

.........tap position of the tap changing transformer in the line km

.........reactive power generation of ith generator

. phase angle of kth bus

. .......conductance of line km

........susceptance of line km
 ........real power generation of kth bus

........reactive power generation of kth bus;

........reactive power load at kth bus

The solution of this problem requires the creation of the Lagrangian as shown below:

................................................................................................(4.65)
Where,

objective function

equality constraints (eq. 4.57 and 4.58)

inequality constraints (eq. 4.59, 4.60, 4.61, 4.62, 4.63, 4.64)

and

- -----vectors of Lagrange multipliers

x ----------state vector

Interpretation of Lagrange Multipliers

Any optimization problem will have a Lagrange multiplier associated with each equality constraint. In
the case of OPF, the Lagrange multiplier associated with each constraint represented by the power
flow equations is the derivative of the total cost with respect to the increase in the load at that bus.
This derivative can then be looked at as the instantaneous price of next small increment of load - or
simply the zone price in INR/MWh.

The scarcity of transmission capacity is reflected by the Lagrange multipliers associated with equations
4.63 and 4.64. General theory of optimization suggests that the Lagrange multipliers indicate how the
optimal solution changes if the relevant constraint is changed marginally. Non-binding constraint
implies zero value of corresponding Lagrange multiplier. In case of binding constraint, corresponding
Lagrange multiplier indicates change in the optimal solution if the constraint was marginally eased.
The Lagrange multiplier associated with line flow limits thus sets a threshold on the price per unit,
which one would be willing to pay in order to increase an available capacity marginally. This threshold
is termed as shadow price. As the transmission resources go on becoming scarcer, the line shadow
price gets reflected in the nodal price as it is the sum of the price for generation and transmission.

Pricing system based on nodal prices is an essential feature of Standard Market Design (SMD)
proposed by Federal Electricity Regulatory Commission (FERC) in USA. The nodal price is popularly
known as Locational Marginal Price (LMP). If there are no lines that are congested, then the nodal LMP
for all buses will be equal if DC OPF is employed. In a full AC OPF, even the uncongested case will
have different bus prices. This is due to the effects of transmission losses. The difference between bus
prices equals the value of marginal losses between the nodes.

The physical meaning associated with LMP can be stated like this: It is the rate of increase of optimum
generating cost with respect to the increase of real power load at that bus. It is the marginal cost of
supplying next 1 MW increment at the corresponding node. Same can be established in mathematical
terms as:

..................................................................(4.66)

where, is the optimal operating cost with the variable load at bus k.

Illustrative Example

For illustration purpose, the Optimal Power Flow problem for a 3 bus system of Figure 4.2 is solved.
System data given in Tables 4.2 and 4.3 is used. The cost curves for generators at bus 1 and 2 are
given as:

In addition, following inequalities are imposed

AC OPF, as explained in section 4.6.3, is solved with the above mentioned constraints with the
objective of generation cost minimization. The results are shown in Table 4.4.

(INR
Bus
No
V PG PD /
MWh)
1 1.099 0 359.87 - 1731.5

2 1.100 10.041 540.13 100 1731.5

3 1.099 -0.893 - 800 1731.5

Table 4.4: Results of OPF without line limits

It is observed that power flow between buses 1 and 2 is 210.85 MW. The value is same for every
bus as no line limits are hit and the system marginal price is same on all nodes. Now, in addition to
the above constraints, real power flow over line 1-2 is limited to 200 MW such that,

By placing a limit of 200 MW real power on line 1-2, a congestion is created, due to which the bus
incremental costs are different at different buses. The Lagrange multiplier associated with the line flow
constraint of line 1-2 comes out to be greater than zero in this case. The results are shown in Table
4.5.

Bus (INR /
No
V PG PD
MWh)
1 1.100 0 382.11 - 1838.22

2 1.100 9.492 517.89 100 1660.35

-
3 1.099 0 800 1832.65
0.909

Table 4.5: Results of OPF with line limits

Implications of Nodal Pricing

As seen from the above example, the nodal prices come out to be different when at least one
Lagrange multiplier associated with line flow constraints is non-zero. The difference in nodal prices
gives rise to congestion charge or the network revenue (NR), which is given as follows

...............................................................................(4.67)

where, ,k and are the nodal price, demand and generation at node k, respectively. For the above
example, during the uncongested case, NR is zero as all nodal prices are
equal. However, in the second case, i.e., when line 2-3 has hit its limit, NR is given as follows:

.....................................................(4.68)
This amount is collected by ISO as it buys power at some nodal price and sells it at buses having
relatively large nodal prices. The ISO may use this excess money for network reinforcements or
distribute it among the participants using the financial transmission rights (FTRs).

A thorough treatment to the topic of interpretation of Lagrange multipliers associated with OPF
problem, LMP calculation and FTRs is provided in the next chapter. The above discussion on this topic
was provided in its most simple form so as to make the reader aware of the concept of nodal pricing
and associated congestion management.

NODAL PRICING: OPF BASED CONGESTION MANAGEMENT

Though the nodal pricing method is supposed to be economically most efficient (ability to maximize
social welfare), it suffers from the limitation of complexity that is associated with it. Zonal aggregation
has been proposed and adopted in some systems as a realistic alternative to nodal prices. The basic
idea is to divide the grid into few congestion zones. When congestion is present, the zonal markets are
decoupled and the zonal market clearing prices reflect the supply and demand condition in each zone
as well as inter zonal transmission capability.

In some systems, there exist only a few lines where frequent congestion occurs. The rest of the lines
are less likely to be congested. This practicality has given birth to the concept of inter-zonal and intra-
zonal congestion management. The main objective is to reduce the complexity of readjustment
process with a large power network, and to avoid price deviation within a portion of the system where
congestion is less frequent. In this concept, the entire network is divided into multiple physical zones
across the potentially congested lines. These lines are considered to be economically significant.

At the first stage, an economically significant readjustment and pricing mechanism is utilized in order
to remove congestion on the potentially congested inter-zonal line. This stage is called inter-zonal
congestion management. In order to facilitate the readjustment, the participants submit the
adjustment bids. The adjustment bid submitted by a participant reflects its willingness to allow
adjustment over its preferred schedule. The adjustment bid consists of offer price, maximum
decrement limit and maximum increment limit. If a participant does not submit the adjustment bids,
the ISO will use adjustment bids of other participants for congestion management and the participant
who did not submit the adjustment bid would be automatically forced to pay congestion charges
calculated according to other bids.

The next stage is intra-zonal congestion management. During intra-zonal congestion management
stage, the flows on the inter-zonal channels are set fixed to the values which are the outcomes of the
inter-zonal congestion management stage. The main goal is to minimize the absolute MW of re-
dispatch by taking into account the net cost of re-dispatch, determined by the adjustment bids of the
participants. Inter-zonal and intra-zonal congestion management scheme is employed in ERCOT
market.

Generic Formulation

In the ERCOT Zonal Congestion Management Market, Qualified Schedule Entities (QSEs) , (a term
used for Scheduling Coordinators in ERCOT) submit balanced energy schedules that include physical
schedules, bilateral transactions, and Balancing Energy bids. QSEs balance their schedules through
bilateral transactions. After conducting an evaluation of congestion conditions using generation
schedules submitted by QSEs and short-term load forecast by ERCOT, ERCOT purchases Balancing
Energy up bids and Balancing Energy down bids from different zones to resolve zonal congestions
while maintaining the balance between generation and load. To minimize system cost and maximum
system welfare, the following objective function is used to calculate zonal Market Clearing Price,
Shadow Price of Zonal Congestion, and Balancing Energy bid deployment.

QSE is an entity that coordinates with ISO on behalf of generators, customers and retailers. It supplies
balanced schedules of power injection and take-offs on hourly basis.

............................................................(4.69)

...............................................................(4.70)

..............................................................(4.71)

..................................................................(4.72)
 .....................................................................(4.73)
Where,

set of QSEs

set of zones
KC .set of zonal congestions
NP total number of QSEs
NZ total number of zones
NC total number of zonal congestions
zonal balancing energy up bid price
BEUBP
curve
zonal balancing energy down bid
BEDBP
price curve
INSU deployed balancing energy up bids
deployed balancing energy down
INSD
bids
QSEs total zonal generation
P
schedules
LF short-term load forecast in zone
SF zonal average shift factor
TTC total transfer capacity
maximum amount of balancing
MAXBEU
energy up bids
maximum amount of balancing
MAXBED
energy down bids

Objective function of equation 4.69 indicates maximization of social welfare. Equation 4.70 indicates
overall power balance. Inequality of 4.71 indicates the limit on total transfer capability of a congestion
corridor. A feasible solution for this optimization function consists of three parts: the amount of
cleared Balancing Energy bids, the Shadow Price of Zonal Congestion, and the Shadow Price of Power
Balance. The Marginal Price, i.e., the Market Clearing Price (MCPE), for each zone is calculated as
follows:

..............................................................(4.74)
Where,

........Zonal market clearing price

....Shadow price of power balance

SP ...............Shadow price of zonal congestion
The deployed Balancing Energy up bids will be paid at zonal MCPE while the deployed Balancing
Energy Down bids will pay ERCOT at the zonal MCPE. Those QSEs whose schedules aggravate a Zonal
Congestion will be charged at the SP of that Zonal Congestion in proportion to their scheduled flow on
that Zonal Congestion. Those QSEs whose schedules contribute to resolving a Zonal Congestion will be
paid in proportion to their scheduled counter flow.

Illustrative Example
Inter-zonal Congestion Management

The concept of inter-zonal/ intra-zonal congestion management can be explained with the help of a
simple illustrative example. Figure 4.3 shows a simple 3 bus, 2 zone system in a certain hour, with
two scheduling coordinators (QSEs).

Figure 4.3: Two zone example

Variables refers to jth generator of ith QSE in zone k. Similarly, variables refers to jth load
of ith QSE in zone k. The preferred initial schedules and incremental
decremental bids of the QSE s are given in Table 4.6.

Suppose the line connecting zone 1 and 2 has power carrying capacity of 500 MW. Since the total
generation in zone 1 as per preferred schedule is 900 MW and load in zone being 300 MW, 600 MW
will flow on line connecting zones 1 and 2, resulting in a 100 MW violation. This is the case of inter-
zonal transmission congestion management. The ISOs congestion relief protocol requires that the
balanced schedule of each QSE be preserved (balanced portfolio). In other words, an increase (or
decrease) in a certain QSEs portfolio is compensated by corresponding decrease (or increase) from
the same QSE.

Thus, decrease in generation of QSE1 in zone 1 with corresponding increase in generation of QSE 1 in
zone 2 (to balance the portfolio), the cost incurred would be (2000-1000=1000 INR/MWh). Thus, QSE
1 places an implicit bid of INR 1000/MWh. Similarly, QSE 2 places an implicit bid of INR 2000/MWh.
Since, bid of QSE 2 is higher than QSE 1, scheduled flow of 300 MW of QSE 2 will not be altered, while
scheduled flow of QSE 1 will be decreased until the line limit is not violated. Thus, scheduled flow of
QSE 1 will be made 200 MW by making generation of equal to 400 MW and that of equal to 100 MW.
Due to requirement of balancing the portfolio, following conditions are satisfied:

.......................................................................(4.75)

........................................................................(4.76)

Pref. Inc/Dec
QSE Variable
Schedule Bid
500 1000

0 2000

QSE1 100 4000

200 -

400 -

400 3000

0 5000
QSE2
100 -

300 -

Table 4.6: Preferred initial schedules and Inc/Dec bids

Congestion Charges: Zone 2 LMP is set by and it is equal to INR 2000/MWh. Similarly, Zone 1 LMP is
set by and equals INR 1000/MWh. Congestion charges for QSE 1 are calculated as follows:

...................................................(4.77)
Similarly, congestion charges for SC2 are calculated as follows:
 ...................................................(4.78)
The ISO thus receives INR as total congestion charges and then allocates the money to transmission
right holders on that path.

Intra-zonal Congestion Management

Figure 4.4 shows zone 2 with nodal injections and line flows after inter zonal congestion relief.

Figure 4.4: Intra-zonal congestion management

It is seen that after the phase of inter zonal congestion management, the flow on line connecting
nodes A and B is 300 MW, whereas, its limit is 250 MW. In order to relieve this intra zonal congestion,
the generator with highest decremental bid on bus A is called for decreasing its generation. In this
particular example, has the highest decremental bid (INR 4000/MWh). Hence, its output is reduced by
50 MW, while generator output will be increased by 50 MW, leading to flow on line AB as 250 MW.

Intra zonal congestion settlement:

Generator belonging to QSE 1 decreased its output by 50 MW. Thus, payment by QSE 1 to ISO is

...............................................................................(4.79)

Similarly, generator belonging to QSE 2 increased its output by 50 MW. Thus, payment by ISO to
QSE 2 is given by

.............................................................................(4.80)
Thus, total balance with ISO is INR. This will be recovered from QSE s as a zonal uplift based on their
load in zone 2. If ISOs balance for intra zonal congestion management is positive, then this uplift is
credited to QSE s based on their load in that zone. In case of negative balance, the QSE s are debited.
In the above case, QSE s will be charged as follows:

QSE 1 pays
...........................................................................(4.81)

QSE 2 pays
.........................................................................(4.82)

Price Area Congestion Management

This is a simplified version of the inter-zonal and intra-zonal congestion management
scheme. This method consists of splitting a power exchange into geographical bid areas with
limited capacities of exchange. When congestion is predicted, the system operator declares
that the system is split into areas at predicted congestion bottlenecks. Spot market bidders
must submit separate bids for each price area in which they have generation or load. If no
congestion occurs during market settlements, the market will settle at one price, which will
be the same as if no price areas existed. If congestion does occur, price areas are
separately settled at prices that satisfy transmission constraints. Areas with excess cheaper
generation will have lower prices, and areas with excess load - higher prices. Market income
from this price difference is paid to the SO and he further uses it for grid enhancement.
Bilateral contracts that span price areas must purchase the loads energy in its price area in
order to account for the contribution to congestion and to expose the contract to the
financial consequences of congestion.

An advantage about the market splitting method is that on a long-term basis, new gencos
may decide to add capacity in deficit zones, attracted by high sale prices, and thus
introduce more competition and cause overall prices to decrease. A limitation associated
with this type of system is that it can be used only when physical zones are connected in
radial fashion. In a meshed system, clear cut boundaries of physical zones can not be
established. This type of congestion management system is used in Norway.

Algorithm for Price Area Congestion Management Scheme

Resolving congestion management by market splitting is a step by step process. Algorithm

1 gives a step by step procedure for price area congestion management. The scheme is
explained with a simple two bus system shown in Figure 4.5.
 Figure 4.5: Two area system

With no congestion, the market will settle at a single unconstrained market price MPU, and
the total generation and load will be equal, i.e.,

.................................................................................(4.83)

Where,

generation in areas A and B respectively

loads in areas A and B respectively

Since price area A has overall cheaper generation, area A will be net exporter and flow on
line AB will be from A to B. If this unconstrained market settlement results in exceeding the
transfer limit of line AB, then each zone is treated as a separate area, and the zonal
markets are resolved separately. The power balance constraint in zone A is that generation
equals area A load plus transfer on line AB. The value of transfer is the maximum transfer
limit of line AB. An example of the supply and demand curves for two areas A and B is
shown in Figure 4.6.

Figure 4.6: Price area settlement for Two zone system

Here, the unconstrained market clearing price is MPU. The exchange capacity is included in
the area price calculation by moving the supply curve in area B and the demand curve in
area A, as indicated by the dotted curves. Now,

.......................................................................................(4.84)
Similarly,

........................................................................................(4.85)

Individual settlements for each market is done once again. This arrangement ensures that
PAB is restricted to . It can be seen from the Figure 5.6 that the area A price (MPA) is lower
than unconstrained market clearing price (MPU), where as the price of area B (MPB) is
higher than (MPU). The difference between the respective area prices and the system price
is called the network rent and is given as:

.....................................................................................(4.86)

Algorithm for Price Area Congestion Management

1. Solve unconstrained market clearing process. That means, the market clearing for
the whole system is done using only bid data and without network constraints.
2. A power flow simulation is carried out using nodal injection data obtained after step
1.
3. Check whether any of the line limit is getting violated.
4. Depending on the congestion bottleneck, form the price areas across the congested
corridor.
5. Do separate market settlements in each price area. Thus, two separate market
clearing prices are formed.
6. Adjust demand curve in cheap price area by giving it a lateral shift to the right.
Similarly adjust supply curve in high price area by giving it a lateral shift to the
right. The amount of lateral shift is equal to the line flow limit of congested
bottleneck.
7. After making adjustments given in step 6 and doing market clearing for individual
areas, it is ensured that transmission congestion is relieved. This leads to two
different prices in two price areas. The settlements are done using respective area
prices.

Illustrative Example

Consider a two area system as shown in Figure 4.5. Suppose the capacity of line between
the system A and B is 100MW. The Power exchange receives buy and sell bids from both
areas A and B. These bids present power as a function of price. These are represented as:

...............................................................................(4.87)

................................................................................(4.88)
The bid constants are given in Table 4.7.
 sell Bid Buy Bid
Area
c d a b

A 15 120 -4 -350

B 8 160 -6 -525

Table 4.7: Bid Data

Step 1: Unconstrained market settlement is done initially. This gives a market price of 35
INR/MWh. The powers in each area are shown in the Table 4.8:

Area PG PD

A 405 210

B 120 315

Table 4.8: Results of unconstrained market settlement

Step 2: Calculate the power flow on line AB. In this example,

.........................................................................(4.89)
Step 3: The results of step 1 are compared with the line flow limits to see if congestion is
occurring on any of the corridors. In this case, the flow on line connecting areas A and B is
195 MW as against its capacity of 100 MW.

Step 4: Separate the market into two price areas: area A and area B.

Step 5: Find out market clearing prices of individual price areas, using load and generation
bid data of respective areas. The outcome of this is presented in Table 4.9.

Area
Area PG PD
Price
A 250 250 24.73

B 231 231 48.95

Table 4.9: Results of individual market settlement

Step 6: Using the principle of market-splitting, each area is now dispatched separately and
by including the line capacity constraint. The cumulative load curve in area A and
cumulative supply curve in area B are adjusted such that flow over a line connecting areas A
and B is restricted to 100 MW. This adjustment renders following functions for individual
market settlements:

..........................................................................(4.90)

............................................................................(4.91)
Step 7: The results of final dispatch are given in Table 4.10.

Area
Area PG PG PAB
Price

100
A 330 230 (net 30.0
export)

-100
B 174.3 274.3 (net 41.78
export)

Table 4.10: Results of constrained market settlement

In this case, after making settlements, there is some money left with the power exchange.
It is equal to differences of area prices times the flow on tie line connecting two areas.

Network_Rent
.............................................................................(4.92)

It is apportioned to the TSOs involved for grid capacity enhancement. Bilateral contracts
that span price areas must sell the energy in the supply area and purchase energy in the
load area to account for the contribution to congestion, and to expose the contracts to the
financial consequences of congestion. It is the only instance of mandatory spot market
participation.

CAPACITY ALLEVIATION METHODS

The methods discussed previously belong to the class of capacity allocation methods which
aim at ex-ante capacity allocation in the day-ahead stage or in general, at the time of fixing
the operation schedules. In spite of this, the possibility of unscheduled interchanges can
never be ruled out and the system operator relies on the defence mechanism employed
during real time operation, in order to relieve congestion. The other class of methods, i.e.
congestion alleviation methods are often used to relieve congestion in real time and are also
referred to as remedial actions. Some of the common congestion alleviation methods are
discussed below.

Re-dispatching

Re-dispatching is exercised as a command and control scheme, i.e., ISO curtails or

increases injections without market based incentives. As generators have to be reimbursed,
the ISO has an incentive to keep re-dispatch cost low.

Counter Trade

Counter trading is based on the same principles as re-dispatching, however, it may be

considered market oriented. Rather than applying command and control, the ISO will buy
and sell electricity at prices determined by a bidding process. The principle of counter
trading is thus a buy back principle which consists of replacing the generation of one
generator ill- placed on the grid as regards to congestion by the generation of a better
placed generator. The ISO has to buy electricity downstream of the congestion at higher
cost and sell it upstream. Thus, there is no congestion rent, instead a congestion cost for
ISO. This cost exposure is also regarded as an incentive for investment into grid capacity.
Counter trading is used for real time congestion relief in the Norwegian system.

Curtailment

As mentioned earlier, a transaction-based curtailment approach is another methodology

that is used for congestion management. An example is NERCs TLR procedure
(Transmission Loading Relief). This method is used by PJM as a last resort after its
alternative congestion management (based on LMPs as discussed earlier). In real-time
operation, the ISO monitors the system for possible security violations. In the event of such
violations occurring or being imminent, the TLR method of curtailing transactions is
exercised. The transactions are prioritized for curtailment on the basis of criteria that
consider the size of the transaction, its relative impact on the congested line flows, and the
firmness level that was fixed before dispatch. We have seen that PTDFs can be used to
determine the incremental impact of any system injection (or any bilateral transaction) on
any line flow of interest. This method of congestion management is more reliability-driven
than market-driven. In this scheme, price and the actual value of transmission are not
important considerations. So, whereas this method gains reliability-wise, it might lose on
the economic front.

COMPARISON AND CONCLUSION

The rescheduling of generation and load resources and corresponding pricing have been
added with a new dimension to the operation of power market as well as system operation
after deregulation. The solution methodology for this problem is referred to as congestion
management. This chapter brings out various congestion management schemes that are
employed in various markets across the world. The methods can be coarsely classified as
market based and non-market based methods. In the market based methods, the system
operator is bothered about the economical efficiency of the generation re-adjustment
process. On the other hand the non-market based methods are driven by the simplicity and
ease of procedure.

The congestion management techniques are strongly coupled to the overall market design.
The nodal pricing or OPF based schemes require a centralized dispatch. In this scheme, the
ISO schedules generators and loads based on their bid data such that social welfare is
maximized. While doing so, network constraints are added to the optimization problem so
that the most economic nodal injection pattern is obtained without creating congestion. The
physical bilateral transaction is scheduled as if the generation injection submitted a zero bid
and the load submitted an infinite bid. Such a transaction is then subject to an ex-post
charge that equals the opportunity cost of transaction, i.e., the cost difference of selling the
power to the pool at the injection node price and buying it back at the take-off node price.
Bilateral traders then opt for buying financial instruments to hedge the price difference risk.
Bilateral traders can also manage transmission price risk by actively participating in
congestion relief by providing incremental/ decremental bids. The nodal pricing scheme
achieves economical efficiency, however, at the cost of complexity and transparency. The
scheme is commonly employed in LMP markets which work according to standard market
design in USA .

The inter-zonal/ intra-zonal scheme tries to strike a balance between the complexity and
economic significance. The basic idea in this approach is to divide the grid into few
predefined congestion zones, which have separate markets whose respective market
clearing prices set the uniform price within the zone. It makes calculation for larger systems
simpler. In this scheme, different objectives are used to correct for inter-zonal and intra-
zonal line loadings. When inter-zonal congestion occurs, bilateral transactions across zones
are subject to ex-post congestion fee based on congestion relief cost between the zones.
Zonal pricing schemes were employed in Californian market.

Price area congestion management scheme used in Nordic pool forms a special version of
zonal congestion management scheme. It divides the system into different price areas when
there is congestion. These areas are divided at congestion bottlenecks. Each area will have
its own market. Price areas are defined pragmatically, based on experience and engineering
judgment. Some additional criteria for analytical determination of price areas is required if
the system is not radial inter-connected [1].

In contrast with the nodal pricing scheme, where implicit auctioning of transmission capacity
is done, the transmission facilities in European interconnections are allocated by means of
explicit auctioning. This separates transmission and energy markets. The mechanisms to
account for the effect of loop flows, a scheme called coordinated auctioning is being
proposed, that requires centralized auctioning of transmission interconnections.

The term - congestion management generally refers to the capacity allocation to various
participants before finalizing the actual schedules of nodal injections. It is possible that the
commonly employed mechanism may not lead to a feasible solution. The last resort
available with system operator is the forceful curtailment of some of the transactions, based
on some criteria. Again, the real time unscheduled flows may hit the line flow limits. For
this, congestion alleviation methods are employed which may take the form of market
based or non market based solution.

Presently congestion management is a challenging aspect in the deregulated market.

Implementation of a particular scheme in a system is driven by historical developments,
demographic parameters, network topology, political openness and so on. Table 4.11 shows
comparison of practical congestion management schemes.

Market
Method Characteristic Allocation Alleviation Example
Based?
Explicit Decentralized
auctioning of European
Yes Yes No
transmission interconnection
Auctioning capacity

Requires
centralized
PJM, New
Nodal dispatch,
Yes Yes No England, New
Pricing implemented
York
in pool based
markets

Can be used
with
Zonal centralized Australia and
Yes Yes No
Pricing dispatch or Nordic pool
using market
splitting

Re-
- Yes No Yes -
dispatch

Replacement
of ill placed
Counter
producer with Yes No Yes Sweden
Trade
better placed
producer

Some norm of
Most of the
Pro-rata allocation, not
No No Yes developing
methods necessarily
countries
economically

Capacity Allocation :Comparison of congestion management methods

 Module 5:Locational Marginal Prices (LMPs)
INTRODUCTION

The Locational Marginal Pricing (LMP) mechanism is one of the most commonly employed tools for
market settlement in the deregulated power system environment. The Locational Marginal Price (LMP)
at a bus signifies the cost of supplying the next increment of load at that bus. The LMP is the sum of
supplying energy marginal cost, cost of losses due to the increment and transmission congestion cost,
if any, arising from the increment. and congestion, if any, arising from that increment. The LMP is the
true indicator of marginal pricing of energy. The calculation of LMPs implicitly involves congestion
management. Compared to other approaches of congestion management, the LMP approach has found
very wide acceptance throughout the world due to its inherent efficiency in the network capacity
allocation.

Many of the successfully running power markets like PJM, NYISO, ISO-NE, CAISO, ERCOT, MISO and
NEMCO have already implemented LMP mechanism in their systems, whereas other markets are now
evolving towards locational marginal pricing. The LMP mechanism was first invented by Dr. William
Hogan in 1992 [1], and introduced at Pennsylvania-New Jersey-Maryland (PJM) ISO. However, the
basis of the LMP mechanism is the theory of spot pricing proposed by Schweppe et al.[2]. The
distinguished feature of the LMP mechanism is that the entire course of power scheduling (pool as well
as bilateral transactions) is done centrally, recognizing system conditions and constraints arising
thereof. The underlying principle of locational marginal pricing is that the energy price varies from one
location to another location in the presence of congestion and loss in the system.

FUNDAMENTALS OF LOCATIONAL MARGINAL PRICING

Locational marginal pricing is a centralized process of market clearing, where it is the responsibility of
the Independent System Operator (ISO) to determine the power dispatch schedules as well as the
energy prices. Unlike system uniform pricing (i.e., unconstrained bidding) approach, network limits
have to be considered while scheduling generators, loads and bilateral transactions. Since network
constraints are considered in the market clearing process, it is not possible to determine the market
equilibrium simply by the intersection of a cumulative supply curve and a cumulative demand curve.
Instead, the power dispatch schedules and energy prices are calculated through an optimization
approach consisting of network and power flow related constraints.

An LMP market may be a single settlement or two settlement market. In case of a single settlement
market, scheduling is done only in day-ahead, whereas both day-ahead and real-time scheduling is
done for a two settlement market. A real-time market is essentially within the hour market. The real-
time scheduling and settlement are further done in different time blocks. For example, in PJM, real-
time scheduling is done in 5- minute time blocks. The real-time scheduling starts with the state-
estimation solution at the beginning of each time block. The state-estimation solution gives the actual
injection by each generator and actual withdrawal by each load at the current point of time. It should
be noted that the above time frames of settlement are not strict and change depending upon the
market needs.

Apart from the offers from generators and bids from the loads, bids are also invited from the bilateral
transactions under locational marginal pricing. These bids are called up to network usage charge bids.
An up to network usage charge bid indicates the price that the concerned entity is ready to pay at
maximum for the scheduled quantity of its transaction. Similar to a load bid curve, a transaction bid
curve is also downward sloping. Apart from bids and offers, self-scheduling is also allowed in some
markets like PJM. Self-scheduling means showing price-unresponsiveness or price-insensitivity. A self-
scheduled generation, load or transaction is alternatively called as an inelastic generation, load or
transaction, respectively. The power dispatch schedules are obtained by maximizing social welfare
function within the network limits. The energy price at a location is defined as the rate decrease of
optimal social welfare with respect to the fixed load increase at that location. These prices are the
locational marginal prices. Generators are paid and loads have to pay the LMPs at the corresponding
locations. For a bilateral transaction, the market player has to pay a network usage charge depending
upon the LMP difference between its sink and source locations.

The bid and offer curves submitted by market participants may be piecewise or purely linear curves.
In case all the bid and offer curves are purely linear, the social welfare function W(P) can be
formulated as follows:

...........................................................................(1)

where,
...................................................................................................................................(2)

d......................Number of load bids

g......................Number of generation offers
.b.....................Number of transaction bids
Pd......................Column vector containing load bid variables
Pg......................Column vector containing generation offer variables
Pb......................Column vector containing transaction bid variables
........... Bid price as a function of quantity for the kth load bid
........... Offer price as a function of quantity for the kth generation offer
............ Bid price as a function of quantity for the kth transaction bid

For zero slope bid and offer curves, the expression of welfare function is simply reduced to the
following,

.....................................................................................................................(3)

In the absence of price-sensitive loads and transaction, the objective function can be alternatively
defined as minimization of generation cost C(Pg), where,

.......................................................................................................................(4)

As before, for zero slope offer curves, the expression of generation cost is reduced to the following:
 .....................................................................................................................................(5)

It should be noted that when the objective function of the market clearing problem is defined as
minimization of generation cost, the definition of LMP is also changed as the rate of increase of
optimal generation cost with respect to the fixed load increase at a location.

Now, consider the case that the generators, loads and transactions submit piecewise linear offer and
bid curves, respectively. Under such a situation, the social welfare function is to be formed by
considering each linear segment as a separate offer or bid. For example, consider the generation offer
curve shown in Fig. 1. This offer curve is equivalent to the set of three individual offers as following:

Figure 1: Generation offer curve

Offer 1: Offered amount = 70 MW,

Offer 2: Offered amount = 70 MW,
Offer 3: Offered amount = 50 MW,

It should be noted that from the point of view of scheduling, the social welfare function is nothing
more than a simple mathematical function that has been devised to reach efficient market equilibrium.
Market equilibrium is said to be efficient if the following conditions hold.

Each selected generation is paid at least its offer price.

The offer price associated with an unselected generation is higher than the energy price at
corresponding location.
 Each selected load or transaction has to pay at most its bid price.
The bid price associated with an unselected load is lower than the energy price at
corresponding location.
The bid price associated with an unselected transaction is lower than the LMP difference
between the corresponding locations.
It is not possible to increase the output of a generator at a certain location, and
simultaneously increase the consumptions at some other location/s without causing fall of at
least one load node energy price below the generation node energy price or creating network
infeasibility, and vice versa.
It is not possible to increase the quantity of a transaction without causing network infeasibility
or making its network usage charge negative.

ILLUSTRATIVE EXAMPLE

First five of the above conditions are explained with an example. Consider that the offer price
associated with a generation offer is $20/MWh and the offer gets fully selected in the energy market.
In case the market equilibrium is efficient, the LMP at this generators location can never be less than
$20/MWh. Similarly, let the offer price associated with a un-selected generation- offer be $30/MWh.
Under efficient market equilibrium, the LMP at the corresponding location must be less than $30/MWh.
The opposite conditions hold for a load or transaction bid.

For the illustration of sixth and seventh conditions for efficient market equilibrium, consider the
sample system shown in Fig. 2. Fig. 3 shows one possible equilibrium state satisfying the other
necessary efficiency conditions. It can be seen that it is still possible to increase the output of the
generator at Node 1 and the consumption at Node 2 with neither causing network infeasibility nor
making the price at Node 2 to fall below the price at Node 1. Therefore, this market equilibrium is
inefficient. The efficient equilibrium state is shown in Fig. 4. The condition on bilateral transaction
scheduling can be illustrated in a similar way.

Figure 2: Sample system

 Figure 3: Inefficient market equilibrium

Figure 4: Efficient market equilibrium

A generation offer or load-bid may be real or virtual. A real load bid or generation offer is always
backed up by a real load or generation asset, respectively, whereas there is no supporting physical
load or generation asset, respectively, for a virtual bid or offer. For example, an entity can submit a
generation offer at a location where it does not have any generating asset. Virtual bids and offers are
allowed only under two-settlement market structure. Any virtual generation or load accepted in a day-
ahead market must be counter-balanced by a same sized load or generation, respectively, in the real
time market. The motivations behind encouraging virtual bids and offers are to quickly demolish the
difference between day-ahead and real-time prices, and to increase the market competitiveness.

ZONAL AND NODAL LMP

Locational marginal pricing is done either on nodal basis or zonal basis [3]. For a nodal market, a
location is defined by a node or a load zone or a generation hub and the impact of each nodal injection
on line flows is individually considered in the power dispatch problem. The dispatch solution has to
respect the flow limit of each line. As far as the definitions of load zone and generation hub are
concerned, they are fundamentally similar. A load zone or a generation hub is basically defined by a
set of nodes and a load or generation distribution vector, respectively. In order to calculate the line
flows, the total load on a load zone is distributed over its constituent nodes by means of the
distribution vector assigned to it. The total generation over a hub is distributed in the same fashion.
However, for a load zone, the total load on it is also same as the sum of the total load on each of its
constituent nodes. This is not true for a generation hub. Additionally, the scheduling and metering for
a load zone is always done in the aggregation level by assuming that the total load on this load zone
naturally remains distributed in a fixed ratio. In case of a generation hub, the scheduling and
metering, in the end, is always done at the nodal level. It is the responsibility of the entity, requesting
a transaction from this hub, to maintain the injection pattern specified by the generation distribution
vector. Each individual load within a load zone has to pay the load zone price only, whereas each
individual generation within a hub is paid the LMP of the corresponding node. The objectives behind
defining load zones are to reduce the spatial price variation and to ease the financial settlement. On
the contrary, the objective behind defining hubs is to maintain consistency with the previous (i.e.,
before nodal pricing) bilateral contractual agreements.

In case of a zonal market, the system is divided into multiple transmission zones and each of these
transmission zones defines a particular location. A transmission zone is basically a set of nodes and
lines that is modeled like a single node in the dispatch problem. Only the inter-zonal line flow limits
are considered during the power scheduling. The inter-zonal line flows are determined by the zonal
sensitivity factors. The main technical challenge with zonal pricing is to appropriately divide the
system into transmission zones so that zonal sensitivity factors may exist. For this, the branch-node
sensitivity factors relating any inter-zonal line should be almost same for all the nodes constituting a
transmission zone.
ADDITIONAL ASPECTS OF LMP MECHANISM

The advantage of locational marginal pricing is multi kinds. First, this kind of energy pricing can
generate useful price signals to identify suitable locations for building new generators, load centers
and transmission facilities. For example, if the LMP at a certain location continuously shows a higher
value, it gives an initial indication that this location may be a suitable place for building new
generators. At the same time, penalizing each bilateral contract with a location dependent network
usage charge applies a strong financial force on it, making it implicitly care for the congestion in the
transmission network. The pool prices also recognize the system limits. Unlike system pricing
approach, the possibility of administrative transmission load relief (TLR) thus becomes very small. The
need for TLR arises only during very rare occasions when the self-scheduled generations, loads and
transactions load the system beyond its capacity such that no solution of the dispatch problem exists.
Moreover, due to a joint optimization based dispatch calculation, transmission capacities are
allocated in optimal way to the energy market participants as per their needs. The less the bid price of
a generation bid the more the amount of it would get selected. Similarly, the more the bid price of a
load or transaction bid the more the amount of it would get selected. The capacity allocation occurs in
optimal way due to measuring the relative needs of all the market part at the same time by a single
and non-profit making entity. However, this feature is not available in the physical transmission right
methodology.

LMP FORMULATION AND IMPLEMENTATION

Calculation of LMP in a market environment involves solving an optimization problem. LMP is the by-
product of the solution to this optimization problem. In this section, we shall discuss various models
existing in literature to calculate LMP.

The LMP models existing in literature can be broadly categorized into two classes, namely, the ACOPF
model and the DCOPF model. The ACOPF model uses the AC formulation to represent power flow and
thus can accurately represent the line flows and loss. On the contrary, the DCOPF models are based
upon DC power flow approximation and, therefore, are less accurate. The DCOPF model can be
lossless or loss compensated. In case of a lossless model, the network loss is ignored at the time of
LMP calculation. Even though ACOPF provides the most accurate LMP, it is quite complex to model its
constraints, and cannot be used for real time applications. Sometimes, obtaining converged solution
for very large systems becomes difficult. However, DCOPF is not plagued by the convergence problem
due to its simplicity. Solution of DCOPF can be obtained with a single run of Linear Programming (LP)
because of the linearity assumption. However, due to these assumptions, the LMP obtained is not very
accurate. The choice between ACOPF and DCOPF is a trade-off between accuracy and speed of
convergence. Some advantages of DCOPF over ACOPF are as follows:

1. Simplicity
2. Speed of convergence
3. Availability of LMP components

Among the two major classes of LMP models, the ACOPF model is prone to lack financial consistency in
general. On the contrary, financial consistency can be obtained in a DCOPF model. Moreover, without
significant loss of accuracy, the DCOPF models are much simpler that the ACOPF models. As a result
most of the LMP markets employ DCOPF models for the purpose of market clearing. Currently, NYISO
is the only market that employs an ACOPF model for calculating locational marginal prices.
 ACOPF

The ACOPF model involves solving a linear/non-linear objective function with linear and non-linear
equality and in-equality constraints. The objective function can be cost minimization or social welfare
maximization. Unlike DCOPF, it is required to solve the non-linear power flow equations, and hence a
non-linear optimization solver is required to obtain a solution. For large systems, the dimension is
quite high and it is generally performed as an offline study. The mathematical formulation is as
follows:

...................................................................................................................(6)

..........................................................................................(7)

..........................................................................................(8)

............................................................................................(9)

.........................................................................................(10)

..........................................................................................(11)

.........................................................................................(12)

(7) and (8) are the real and reactive power balance equations for all buses in the system. (9) models
the line flow limit in-equality constraints. (10) and (11) enforce generations limits on real and reactive
power for all generators. Voltage limits are enforced using (12). Additional constraints pertaining to
taps and shunts may also be added to the optimization problem. For a practical system, the problem
size is quite large and hence obtaining a converged solution sometimes becomes difficult.

The LMP at a bus is the lagrangian multiplier corresponding to the equality constraint. Unlike in
DCOPF, where the LMP components are readily available, additional de-composition techniques are
required to obtain LMP components [4-11]. Moreover, there exists no unique solution to this
decomposition. Perhaps, this is one the driving factors for DCOPF to be preferred over ACOPF.

DCOPF
Nomenclature

The following is the list of symbols used in the DCOPF models to follow.

n.............................................Number of buses/nodes
M.............................................Number of Transmission Lines
Ci(Pgi).....................................Cost curve of generator located at bus i
g............................................Number of Generator offers
LF............................................Column vector containing loss factors
LF.............................................Matrix containing line-bus loss factors
..............................................Lagrangian multiplier of equality constraint
..............................................Column vector containing Lagrangian multiplier of in-equality
constraints
SF............................................Sensitivity matrix relating Power injections and line flows
L..............................................Lagrangian function
Ploss.........................................Total system power loss
...................................... Power loss occurring on line k
..................................... Column vector containing power loss occurring on all lines
flowmax...................................Column vector containing line power flow limits
Pgi.......................................... Power generation at node i
Pdi..........................................Power demand at node i
Pg............................................Column vector containing Power generation at all buses
Pd ...........................................Column vector containing Power demand at all buses
....................................Lower limit on Pgi
o..................................Upper limit on Pgi
D.............................................Loss distribution vector
K.............................................Loss distribution matrix
FND........................................Column vector of fictitious nodal demand
offset......................................Offset term required to compensate for over-estimation of losses
offsetL............................... ......Column vector containing offset for all lines
1x ..........................................x * 1 vector of all ones
Rk .........................................Resistance of line k
Xk .........................................Reactance of line k
flowk ....................................Power flow on line k

Lossless Model

This is the most simplistic form of DCOPF, where the transmission line losses are completely
neglected. The solution of lossless DCOPF is the starting point for some of the approaches that
incorporate losses since it provides a good initial estimate. The formulation of lossless DCOPF is given
below.

..................................................................................................................(13)

.....................................................................................................................(14)

...............................................................................................(15)

..............................................................................................................(16)

Equation (13) is the objective function. (14) is the global power balance equation. (15) ensures that
the transmission line flows don't exceed the limits. (16) is the constraint on individual generator limits.

The Lagrangian for the above optimization problem can be written as follows.

................................................................(17)
For the sake of simplicity, the flow constraints are considered for only one direction of power flow. To
obtain the LMP at any bus i, the partial derivative of the Lagrangian w.r.t. the demand at that bus i is
evaluated, given as follows:

.....................................................................(18)

As seen from equation (18), the LMP obtained from lossless DCOPF consists of two components. The
first term (?) is referred to as the energy component (?_e), since it is the Lagrangian multiplier of the
global power balance equation. The second term is known as the congestion component (?_c) as it is a
function of the Lagrangian multiplier of the in-equality constraint that corresponds to line power flows.
It must be noted here that in the absence of congestion (=0), the LMP at all buses will turn out to be
same. This holds true only in case of DCOPF. When we use ACOPF, the LMP at all buses will be
different because of the presence of transmission losses, even in the absence of congestion.
The features of LMP are better understood using an example, which is discussed next.

Example 1: A sample 5 bus system given in Figure 5. The system has five dispatchable
generators. The line data for the system is given in Table 1. The marginal costs of all
generators are given in Table 2. The loads are inelastic and are 300 MW each.

Line
A-B A-D A-E B-C C-D D-E
Data

R (%) 0.281 0.304 0.064 0.108 0.297 0.297

X (%) 2.81 3.04 0.64 1.08 2.97 2.97

Limit
999 999 999 999 999 240
(MW)

Table 1: Line data of 5 bus system

 Figure 5: Sample 5 bus system

Marginal
Generator Cost
($/MW)

G1 14

G2 15

G3 30

G4 35

G5 10

Table 2: Marginal costs of generators

The results for the lossless case, when bus 4 is taken as slack, are given in Tables 3-5.

Dispatch
Generator
(MW)

G1 110

G2 100

G3 0

G4 116.0757

G5 10 573.9243
 Table 3: Dispatch for lossless case

Power
Line
flow(MW)

A-B 379.7505 0

A-D 164.1738 0

A-E -333.9243 0

B-C 79.7505 0

C-D -220.2495 0

-
D-E -240
52.0344

Table 4: Power flow for lossless case

LMP Components
($/MW) LMP
Bus
($/MW)
Energy Congestion

A 35 -19.1744 15.8256

B 35 -11.3202 23.6798

C 35 -8.3015 26.6985

D 35 0 35

E 35 -25 10

Table 5: LMP at various buses

The following observations about LMP can be made from the results:

1. The energy component will always be equal to the LMP at the reference/slack bus.
2. The energy component at the reference/slack bus also consists of its congestion component
and loss component (if any). The energy component has been named so just for
representation and does not indicate the cost of supplying only energy.
3. Even if the reference bus is changed, the energy and congestion components will change, but
the total LMP will remain the same.
4. In the absence of congestion, LMP at all buses will be same, and equal to that of the marginal
generator.
5. The availability of LMP components is an inherent feature of DCOPF. D e-composition of LMP
into its components is required for the settlement of risk-hedging tools like financial
transmission right (FTRs) [12]. FTRs are basically risk hedging instruments whose objective is
to guard forward contracts from the uncertain congestion charges.
The lossless model always yields a reference node independent solution. This is one the desired
features of LMP calculation because there shouldn't be inconsistency in financial settlements. Financial
consistency mainly implies non-negativity of net congestion collection or adequacy of net congestion
collection to make full payments towards a set of simultaneously feasible FTRs; whichever is of
interest. Some desirable features of LMP calculation are the financial consistency, reference node
independency, accurate modeling of power flow, and the model simplicity. The reference node
independency of LMP calculation is desired in order to obtain an undisputed solution of each quantity
of interest. Both, reference node dependent and independent loss-compensated models are available
in the literature for DCOPF formulations.

Even though the lossless DCOPF model is simple to formulate, it has certain disadvantages. For large
systems, the losses are not negligible and hence can't be neglected in the dispatch because the
mismatch will be quite large. This has forced researchers to improve upon the lossless DCOPF model
such that losses are considered in the dispatch.

Model 1

In this model, the losses are estimated from pre-calculated loss factors. The lossless DCOPF model is a
good starting point to evaluate loss factors. These loss factors can be evaluated using AC power flow
(ACPF) or DC approximation. The calculation of these loss factors is dealt with in the next section. The
mathematical modelling of this model is as given below.

....................................................................................................................(19)

...............................................................................................................(20)

...............................................................................................................(21)
-flowmaxSF x [Pg - Pd]
max
flow ................................................................................................................(22)
Pgmin Pgi
max
Pg .................................................................................................................(23)

The difference between this model and the earlier model is in considering the system losses as part of
the dispatch. Hence, the total dispatch will account for the losses occurring in the system as well.
However, the loss distribution is not specified in this model. If we look at equation (22) closely, there
is no mention of loss being compensated at individual bus level. This signifies that all losses are being
consumed at the slack/reference node itself, instead of being distributed across the network . Even
though the dispatch accounts for transmission losses, there is no effect on line flows. This is a major
drawback of this model. Further, the solution of the model changes if the slack bus is changed. Hence,
in 2004, Eugene Litvinov et.al. proposed a vector loss distribution (VLD) model that accounts for
losses at individual bus level, which we will discuss next

Model 2 [12]

In this model, the losses are distributed as additional loads to buses using a loss distribution vector
(D). The problem formulation for this model is as follows.
 .................................................................................................................(24)

such that
.................................................................................................................(25)

.................................................................................................................(26)
-flowmax SF x [Pg - Pd -D x Ploss] flowmax
.................................................................................................................(27)
Pgimin Pgi Pgimax.
.................................................................................................................(28)

The loss factors in this model are calculated on the lossless snapshot using the fast-decoupled
approach (a variant of calculating loss factors using ACPF). The difference between this model and
Model 1 can be seen when we compare equation (27) and (22). The D vector assigns the total system
losses as additional load to various buses. This affects the line power flow, which is desired because
losses occur on transmission lines and are spread over the transmission network. The Lagrangian for
this model is as follows.

.
.................................................................................................................(29)

Because of the introduction of the additional constraint that considers losses, the definition of LMP in
this model is different. Apart from the energy and congestion component, an additional component
called loss component (?1) is introduced. The LMP at any bus i can be defined as follows:

.
.................................................................................................................(30)

Ploss is a control variable in the optimization problem. Hence, the derivative of the Lagrangian function
is necessary to obtain the three components of LMP.

.
.................................................................................................................(31)

From the above two equations, LMP can be defined as:

.
.................................................................................................................(32)
Thus, the LMP obtained from DCOPF inherently consists of the individual components. As such, no
separate technique is required to obtain these components. Hence, LMP can be written as a sum of the
three components as follows:
.
.................................................................................................................(33)
. .................................................................................................................(34)
. .................................................................................................................(35)

.
.................................................................................................................(36)

Even with the introduction of the loss distribution vector, this model has certain drawbacks.

1. The choice of D is an open research problem since there can be many solutions. The authors
overcome this drawback by choosing D proportional to the demand at a bus.
2. The slack dependency of Model 1 is converted to Dvector dependency of this model. If both
are chosen to be same, both models produce identical results.

With the inclusion of network loss in the LMP calculation, an additional term called the loss component
is added to the LMP. The LMP decomposition should be made in such a way that the loss component
vanishes when the system is lossless and that the congestion component vanishes when the system
is uncongested. The implication of reference node independency for a lossy model of LMP calculation is
threefold:

1. The power dispatch solution should be independent of the choice of reference node.
2. The locational marginal prices should be independent of the choice of reference node.
3. The congestion components of locational marginal prices should be independent of the choice
of reference node.

Model 3 [16]

In 2007, Fangxing Li proposed an iterative technique that uses DC loss factors to obtain LMP. The
mathematical formulation of this model is given below.

.
.................................................................................................................(37)

such that : .
.................................................................................................................(38)

.
.................................................................................................................(39)
-flowmax SF x [Pg - Pd -FND] flowmax .
.................................................................................................................(40)
Pgimin Pgi Pgimax.
 .................................................................................................................(41)

This model introduces the Fictitious Nodal Demand (FND) vector that distributes losses to different
buses. The losses occurring on a line are halved and added as additional loads to the buses connecting
the line. This is a reasonable assumption and provides good accuracy in terms of power flow. The
expression for FND for a bus i is given in (42).

.................................................................................................................(42)

Where, l is the index of lines connected to bus i. Since this is an iterative technique, the procedure of
calculation of LMP is understood easily using the flowchart given in Figure 6. The loss, offset and FND
are estimated from previous iteration, and passed as parameter to the next iteration. Since this is an
iterative technique, these values will converge and the difference between two iterations will be less
than the tolerance. This is different from the formulation of

Model 2. The Lagrangian function and LMP calculation is done as follows.

.....................................................(43)
The LMP is given by:

...........................................................................(44)
 Figure 6: Flow chart of iterative technique proposed by Fanxing Li

Model 4 [17]

Sarkar et.al proposed a Matrix Loss distribution (MLD) method to distribute losses to various nodes.
This method provides additional design flexibility in distributing losses to more than two nodes and in
different proportions. Its mathematical modelling is as given below.

.......................................................................................................(45)

..................................................................................................(46)

.............................................................................................(47)

............................................................................................(48)
Pgimin Pgi Pgimax.........................................................................................................(49)

The loss occurring on each line is a control variable in this model. Hence, the partial derivative of
Lagrangian function w.r.t these controls will exists and appear in the expression for LMP. This is
similar to the calculation of LMP discussed in Model 2.
Model 5 [18]

In 2010, a truly reference independent iterative model was proposed by Furong Li et.al. This model
obtained loss factors from the converged ACPF solution by making use of Newton Raphson Jacobian
sensitivities. The difference between this model and Model 4 is in the calculation of loss factors. The
solution of LP is passed on to NRLF to obtain the loss factors, offset and D. These are then passed as
parameters to the next iteration. The mathematical model is given below.

................................................................................................................(50)

.................................................................................................................(51)

................................................................................................(52)
-flowmax SF x [Pg - Pd -D x Ploss] flowmax
...........................................................................................(53)
Pgimin Pgi
Pgimax................................................................................................................(54)

In (53), D is calculated using (55).

................................................................................................................(55)
The LMP decomposition of this model is similar to the one discussed in Model 2

Loss Factors

DC loss factors

In some of the methods discussed earlier, loss factors are estimated using DC approximation. Since
voltages are assumed to be 1 p.u., the losses can approximated as:

..............................................................................................................(56)

The loss factor (LF) at a bus i is defined as the partial derivate of the above expression w.r.t. power
injection at bus i(Ploss). Hence,

..............................................................................................................(57)
Transmission line flow can be written as:

..............................................................................................................(58)

LF can be further expanded as:

..............................................................................................................(59)

(59) is the expression for loss factor using DC approximation. The offset is required since product of
LF and Pi over estimates the losses. In [16], it has been proved that this product produces double the
system losses. And hence, the offset needs to be negation of (53), which is the actual loss occurring in
the system.

AC Loss Factors

The loss factors using ACPF makes use of the NRLF Jacobian to obtain sensitivities. Additional
sensitivity of Ploss with load angle (d) and voltage magnitude (V) needs to be obtained.

..............................................................................................................(60)

Where, S is the inverse of NRLF jacobian (J). From the above equation, we can write LF as:

..............................................................................................................(61)

The offset in this case is calculated as follows.

..............................................................................................................(62)
 SUMMARY

Use of LMP for energy pricing can generate useful price signals to identify suitable locations for
building new generators, load centers and transmission facilities. The LMP mechanism promotes the
optimal and efficient use of system resources like generation and transmission. However, even though
the LMP mechanism is economically very efficient, caution must be exercised before its
implementation. The LMP is very volatile as real time approaches and safeguard mechanisms like FTRs
are required to hedge against this uncertainty. A certain level of maturity on the part of the market
and its participants is required for these mechanisms to be effectively implemented and remain in
sync. Not all market participants may understand the intricacies involved in the LMP mechanism. It is
a crucial link to various other mechanisms involved in a Centralized dispatch market environment.

REFERENCES

1. W. W. Hogan, Contract networks for electric power transmission, Journal of

Regulatory Economics , vol. 4, pp. 211-242, 1992.
2. F. C. Scheweppe, M. C. Caramanis, R. D. Tabors, and R. E. Bohn, Spot pricing of
electricity, Norwell, MA: Kluwer, 1988.
3. J. E. Price, Market-based price differentials in zonal and LMP market designs, IEEE
Transactions on Power Systems , vol. 22, no. 4, pp. 1486-1494, November 2007.
4. L. Chen, H. Suzuki, T. Wachi, and Y. Shimura, Components of nodal prices for electric
power systems, IEEE Transactions on Power Systems , vol. 17, no. 1, pp. 41-49,
February 2002.
5. J. D. Finney, H. A. Othman, and W. L. Rutz, Evaluating transmission congestion
constraints in system planning, IEEE Transactions on Power Systems , vol. 12, no. 3,
pp. 1143-1150, August 1997.
6. T. Wu, Z. Alaywan, and A. D. Papalexopoulos, Locational marginal price calculations
using the distributed-slack power-flow formulation, IEEE Transactions on Power
Systems , vol. 20, no. 2, pp. 1188-1190, May 2005.
7. X. Cheng and T. J. Overbye, An energy reference bus independent LMP
decomposition algorithm, IEEE Transactions on Power Systems , vol. 21, no. 3, pp.
1041-1049, August 2006.
8. T. Orfanogianni and G. Gross, A general formulation for LMP evaluation, IEEE
Transactions on Power Systems , vol. 22, no. 3, pp. 1163-1173, August 2007
9. L. Wang and M. Mazumdar, Using a system model to decompose the effects of
influential factors on locational marginal prices, IEEE Transactions on Power
Systems , vol. 22, no. 4, pp. 1456-1465, November 2007.
10. E. Litvinov, T. Zheng, G. Rosenwald, and P. Shamsollahi, Marginal loss modeling in
LMP calculation, IEEE Transactions on Power Systems , vol. 19, no. 2, pp. 880-888,
May 2004.
11. Y. Fu and Z. Li, Different models and properties on LMP calculations, IEEE/Power
Eng. Soc. General Meeting , June 18-22, 2006.
12. E. Litvinov, LMP implementation in New England, in IEEE/Power Eng. Soc. General
Meeting , June 18-22, 2006.
13. Z. Li and H. Daneshi, Some observations on market clearing price and locational
marginal price, IEEE/Power Eng. Soc. General Meeting , June 12-16, 2005.
14. B. C. Lesieutre and I. A. Hiskens, Convexity of the set of feasible injections and
revenue adequacy in FTR markets , IEEE Transactions on Power Systems , vol. 20,
no. 4, pp. 1790-1798, November 2005.
15. G. B. Shrestha and P. A. J. Fonseka, Congestion-driven transmission expansion in
competitive power markets, IEEE Transactions on Power Systems , vol. 19, no. 3, pp.
1658-1665, August 2004.
16. F. Li and R. Bo, DCOPF-based LMP simulation: Algorithm, comparison with ACOPF,
and sensitivity, IEEE Transactions on Power Systems , vol. 22, no. 4, pp. 1475
1485, Nov. 2007.
17. V. Sarkar and S. Khaparde, DCOPF-based marginal loss pricing with enhanced power
flow accuracy by using matrix loss distribution, IEEE Transactions on Power Systems
, vol. 24, no. 3, pp. 1435 1445, Aug. 2009.
18. Z. Hu, H. Cheng, Z. Yan, and F. Li, An iterative LMP calculation method considering
loss distributions, IEEE Trans. Power Syst., vol. 25, no. 3, pp. 1469 1477, Aug.
2010.
19. A. J. Wood and B. F. Wollenberg, Power Generation, Operation and Control. John
Wiley & Sons, 1996.
 Module 6: Ancillary Services
INTRODUCTION

Most of the issues and corresponding debates in deregulated power industry pertain to the socio-
economical aspects of the system and the welfare of the society at large. However, the power system
infrastructure and the physical laws that govern its utilization continue to be the same, be it vertically
integrated utility or the restructured industry. Similarly, the activities of the system operator
pertaining to operation and control of the system existed during vertically integrated era and continue
to exist in the restructured era. These activities basically stem from the responsibility of the system
operator to keep the system in synchronism and operate it reliably. In the restructured environment,
these activities are typically known as ancillary services .

Provision of ancillary services under the deregulated environment is not as straight forward as it is
described in the vertically integrated structure. Though many reasons can be figured out, the main
reason is that the entities providing ancillary services may not be under direct control of the system
operator. This issue is highlighted with following two examples. The generators in the competitive
market are scheduled as per the bids provided by them to the market. A power system that has
generation just enough to support the overall load and losses is still a vulnerable system. The system
should have provision for additional generation during contingencies like generator outages. Total
capacity of some generating units can be partly dispatched for energy and partly kept ready for
reserve. For a particular privately owned generating unit with fast ramp rate, the system operator is
likely to schedule most of its capacity as a reserve. The generator, however, may not agree to this
unless and until some compensation is provided to it for maintaining its capacity as a reserve. Thus,
the development of compensation mechanism for this generator, as well as the cost allocation to
customers in an optimal fashion, poses a challenging problem in the restructured environment.

Imagine another situation where, a particular generator is operating on the boundary of its capability
curve. An action demanded by the system operator to increase the reactive power injection for this
generator requires decrease in real power injection. This action, though essential from the system
security perspective, is against the market decision and again, the generator may not agree to do so
without proper compensation. Both these situations do not create much problem in the vertically
integrated structure because, under that regime, the generating units are directly under the control of
system operator, who also is a part of the vertically integrated utility.

A power system would continue to run in synchronism or be stable, unless and until it is perturbed.
However, a practical power system having thousands of elements like transmission lines,
transformers, switchgear, shunt elements, generators, etc., is certainly prone to disturbances. The
system operator has to keep all state variables within the practical limits under normal as well as
under contingent situation. Moreover, it has to achieve the global load-generation balance at all times.
Thus, even though the power industry has been deregulated, the technical activity of control and
operation remains the same. However, the manner in which some of the facilities are procured or
provided by the system operator may change, depending upon the type of control.

In the vertically integrated era, whole of the resources required to take corrective control actions are
under the control of a single utility. This is not the case under competitive environment. Some of the
resources required for control of the system belong to other entities. Thus, they are not mandated to
act as per the directives of the system operator, unless and until a commercial agreement exists
between them. Thus, these are the services which are required to be purchased by the system
operator on commercial terms. As is apparent, the main commodity that is traded in the power market
is the electrical power or the electrical energy. However, to make the transactions of power exchange
or some other trading platform to be feasible, with security and quality, some supporting services are
required, which are nothing but the ancillary services. The word ancillary stems from the fact that they
support the main activity of the market, i.e., trading of electrical energy or power. However, since the
stability and security is at stake without these services, they certainly have a primary role to play.
Ancillary services are defined as all those activities on the interconnected grid that are necessary to
support the transmission of power while maintaining reliable operation and ensuring the required
degree of quality and safety. It becomes clear that the ancillary services may include scheduling and
dispatch, frequency regulation, voltage control, generation reserves, etc. It is the matter of debate
and market design about how to procure these ancillary services. There are some services which can
be provided competitively and some services which come under the direct control of the system
operator.

Types of Ancillary Services

A large number of activities on the interconnected grid can be termed as ancillary services. During the
process of defining the ancillary services, some proposals tried to define 60 different ancillary services!
In order to remove this large discrepancy, the North American Electric Reliability Council (NREC) along
with Electric Power Research Institute (EPRI) has identified 12 functions as ancillary services. These
are:

1. Regulation: The use of generation or load to maintain minute-to-minute generation-load

balance within the control area.
2. Load Following: This service refers to load-generation balance towards end of a scheduling
period.
3. Energy Imbalance: The use of generation to meet the hour-to-hour and daily variations in
load.
4. Operating Reserve (Spinning): The provision of unloaded generating capacity that is
synchronized to the grid and can immediately respond to correct for generation-load
imbalances, caused by generation and /or transmission outages and that is fully available for
several minutes.
5. Operating Reserve (Supplemental): The provision of generating capacity and curtailable
load to correct for generation-load imbalances, caused by generation and /or transmission
outages, and that is fully available for several minutes. However, unlike spinning reserves,
supplemental reserve is not required to respond immediately.
6. Backup Supply: This service consists of supply guarantee contracted by generators with
other generators or with electrical systems, to ensure they are able to supply their consumers
in case of scheduled or unscheduled unavailability.
7. System Control: This activity can be compared with the functions of the brain in the human
body. System control is all about control area operator functions that schedule generation and
transactions and control generation in real time to maintain generation load balance.
8. Dynamic Scheduling: It includes real-time metering, tele-metering along with computer
software and hardware to virtually transfer some or all of generators output or a customers
load from one control area to another.
9. Reactive Power and Voltage Control Support: The injection or absorption of reactive
power from generators or capacitors to maintain system voltages within required ranges.
10. Real Power Transmission Losses: This service is necessary to compensate for the
difference existing between energy supplied to the network by the generator and the energy
taken from the network by the consumer.
11. Network Stability Services from Generation Sources: Maintenance and use of special
equipment (e.g., PSS, dynamic braking resistances) to maintain secure transmission system.
12. System Black Start Capability: The ability of generating unit to proceed from a shutdown
condition to an operating condition without assistance from the grid and then to energize the
grid to help other units start after a blackout occurs.

It should be noted that identification and definition of a particular ancillary service is system
dependent. There is no global definition of a particular ancillary service that is applicable in all
systems. There can be many other possible definitions or combinations.
 Classification of Ancillary Services
There can be various ways of classifying the above ancillary services. One common approach would be
to identify when and how frequently these services are required by the system operator [23]. Thus,
three groups can be formed:

1. Services required for routine operation:

These are the services which the system operator requires quite frequently. Some of these may be
required to provide corrective action on minute-to-minute basis. Following services can be grouped
under this category:

(a) System control

(b) Reactive power support
(c) Regulation
(d) Load following
(e) Energy imbalance
(f) Real power loss displacement

2. Services required to prevent an outage from becoming a catastrophe:

These services prevent the system from going out of step even if a major event occurs. These do not
come into picture on daily basis, or rather; no proactive measures are required to be taken either by
the system operator or the service provider on daily basis. Their effectiveness is sensed only under
contingent situation. Following services fall under this category:

(a) Spinning reserve

(b)Supplemental reserve
(c)Network stability services

3. Services needed to restore a system after blackout:

Re-energizing the system after complete blackout requires support from certain generating stations,
which can pickup generation even in the absence of external electricity support. Such generating units
provide the system black start capability. These services are very rarely used.

A closer look at the list of ancillary services reveals that they are either related to:

1. generation-load balancing issues or

2. the network security related issues.

In further sections, load-generation balancing related services, as well as reactive power support
services, under the network security domain are discussed in detail.

LOAD GENERATION BALANCING RELATED SERVICES

There can be various ways of classifying the above ancillary services. One common approach would be
to identify when and how frequently these services are required by the system operator [23]. Thus,
three groups can be formed:

Frequency deviations, if large enough, may lead to total system collapse. If the system frequency
drops drastically due to a sudden mismatch of load and generation, the under-frequency protection
relays isolate the generating units to avoid damage. This dis-connection of generating units further
increases the drop in frequency. This unbalance and series of incidences may cause disconnection of
tie lines and affect the stability of neighboring control areas.

The imbalance between load and generation may arise due to uncertainties in demand forecasting,
generators' inability to follow up the changes in load and generation or load trips. The vertically
integrated utility was responsible for maintaining sufficient generation to cope up with the load
variations and maintain the reliability. However, under the restructured environment, many questions
arise about which generator should take up the burden of providing the balance and how it would be
compensated for the same.

Figure 6.1 shows three components of load variations during one scheduling hour. The first element is
the average load during the scheduling period, 65 MW over the one hour shown in this case. The
second element is the trend during the hour. In this case, this element increases from -5 MW at 7 am
to +10 MW at 8 am. The third element is the rapid fluctuations in load around the underlying trend.
Here, the fluctuations range over 2 MW. Combined, the three elements yield a load that ranges from
60 to 75 MW during this hour.

The rapid fluctuations are handled by the regulation service. The trend of slower change during that
hour is taken care by load following service. Load following involves handling the imbalance at the end
of scheduling interval. These two services plus energy imbalance together ensure that, under normal
operating conditions, a control area is able to continuously balance generation to load. Explanation of
two services is provided next.

Figure 6.1: Components of a load

Frequency Regulation

Regulation is the use of online generating units that are equipped with control mechanism that can
change output quickly to track the moment-to-moment fluctuations in load and unintended
fluctuations in generation. In doing so, regulation helps to maintain the system frequency, minimize
the differences between actual and scheduled power flows between control areas, and match
generation to load within the control area. Figure 6.2 shows the simplified block schematic for
frequency regulation.

The frequency regulation consists of two modes of operation, namely, primary regulation and
secondary regulation. The primary regulation is basically the governor action with certain droop
characteristic, while secondary regulation pertains to Automatic Generation Control (AGC). Figure 6.3
shows the drooping governor characteristics of a generator connected to the system. This droop
characteristic is facilitated by a frequency error feedback provided to the governor. The slope of the
characteristic determines the change in the units output for a given change in frequency. This unit
regulation is provided in percentage. For instance, 4% regulation for a unit indicates that a 100%
change in unit output requires 4% change in frequency. This corrective action facilitated by governor
response is termed as primary regulation.

Figure 6.4 shows drooping characteristics of two units connected to a common load. A change in load
in this case, will be shared by both the units, depending on their droop characteristics so as to operate
at a unique frequency. At nominal frequency f0, unit 1 and unit 2 outputs are P1 and P2 respectively.
In case of a load increase, the units are slowed down and the governors increase the output until the
units settle at a new operating frequency f. Each unit shares additional load in proportion to slope of
its droop

characteristic, so that their outputs are and , respectively.

Figure 6.2: Simplified block schematic for frequency regulation

Thus, the generation increase, taking place within few seconds along with support from frequency
dependent loads, arrests any further fall in frequency. The system then (in the absence of proportional
and integral control) stabilizes and operates at a new frequency that is slightly less than the nominal
frequency. This mechanism is sometimes referred to as Free Governor Mode Operation (FGMO).

This whole control action causes unscheduled power flows on tie-lines. To restore the system to
nominal frequency, the generation set point of some units should be readjusted, based on new
generation-load balance. In some instances, this is done through an automatic control action known
as Automatic Generation Control (AGC), and is referred to as secondary control. The secondary loop
basically aims at bringing the Area Control Error (ACE) to zero so that tie line flows between the
control areas are maintained to specified values. In some systems, this is achieved through manual
adjustment of governor set point. This control action is sometimes referred to as tertiary regulation.

Figure 6.3: Generator droop characteristics

Figure 6.4: Two units sharing same load

Load Following

Load following is the use of online generation equipment to track the intra and inter-hour changes in
customer loads. Unlike the minute-to-minute fluctuations, which are generally uncorrelated among
customers, the long-term changes in customer loads are generally correlated with each other.
Spinning Reserve Services

Unlike the regulation and load following services, the reserve services are designed to be activated
during large power deficits under a contingent situation. The frequency plot associated with the case
presented in Figure 6.1 is shown in Figure 6.5. In this, it is assumed that at 7:45 A.M., a big generator
is suddenly disconnected. This is the situation when reserve services should come into play.
Depending upon the minimum time in which the generation should start providing corrective action,
the ancillary services are classified into following two categories:
1. Spinning reserve services
2. Supplemental reserve services

Figure 6.5: Roles of reserve services

The spinning reserves must start responding quickly to the frequency change. These reserves are
available in the form of synchronous generators that are synchronized with the grid and generate at a
level lower than the maximum rated capacity. This response must be available within 10s and should
be sustainable for further 20s. There are various issues of debate related to definition of spinning
reserves. There are system specific peculiarities which are reflected in the definition of spinning
reserve of that particular system. A discussion on this debate is provided in [24]. Thereby, the authors
have provided a general definition of spinning reserves as follows:

The spinning reserve is the unused capacity which can be activated on decision of the system operator
and which is provided by devices which are synchronized to the network and able to affect the active
power.

Thus, the authors shed light on two important issues: whether spinning reserves are activated
manually or are automatic; and whether only generators can provide spinning reserve or demand side
can also participate as a spinning reserve? Above generic definition of spinning reserve states that
they are activated by system operator manually and in stricter sense, the demand side can also
participate as a spinning reserve.

The units providing supplemental services need not start responding immediately. The supplemental
reserve services are provided by the generators that have fast start-ups such as gas or oil fired
generators or hydro generators. This response must be fully available within 30s of the incidence and
must be sustainable for further 30 minutes. As can be seen from Figure 6.5, the spinning reserve
succeeds in arresting the frequency drop before it reaches the statutory limit. The supplemental
reserve then helps to bring the system frequency closer to nominal value.

While providing supplementary services, the ramp up rate of the generating unit decides its ability to
provide the reserve service. However, this is not the only deciding factor. The transmission constraints
can limit the ability of a particular unit to act as a reserve provider.

VOLTAGE CONTROL AND REACTIVE POWER SUPPORT

SERVICE
System operators use reactive power resources to maintain the voltage at all the buses around the
nominal value. Keeping transmission level voltages at nominal value or within a tight range ensures
proper voltages at the distribution levels. Another important factor is that the transmission network
security is closely associated with the voltage profile. Since the voltage on a bus is strongly coupled
with the supply of reactive power, the voltage control service is also called reactive power support
service.It is prudent to control the bus voltages by providing for reactive power locally, rather than
making it to flow through the grid. There are three major reasons for this. First, the power system
equipment is designed to operate within a range of voltages, usually within 5% of the nominal
voltage. At low voltages, the performance of most of the electrical equipments is poor. For example,
induction motors can overheat and get damaged. High voltages not only damage the equipment but
lalso shorten their life.
Second, the power transmission capability available from a transmission line design is limited by
technological as well as economical constraints. The reactive power consumes transmission and
generation capacity. To maximize the amount of real power that can be transferred across a
congested transmission interface, reactive power flows must be minimized. Similarly, reactive power
production can limit a generators real power capability.Third, moving reactive power on the
transmission system incurs real power losses. Thus, additional energy must be supplied to replace
these losses.

Different Sources of Reactive Power

Reactive power support can be provided by active sources like generators and synchronous
condensers, as well as by locally installed passive elements like capacitors or inductors. Power
electronics based family of devices called Flexible AC Transmission System (FACTS) can also act as
reactive power support devices. Various sources of reactive power support have different
characteristics in terms of dynamics and speed of response, ability of voltage changes, capital costs,
operating costs and opportunity costs. Their technical as well as economical considerations are
discussed next.

Generators

The synchronous generators are very fast reactive support devices. The ability of a generator to
provide reactive support depends on its real-power production. Figure 6.6 shows the limits on real and
reactive production for a typical generator. This is also called as a capability curve of a generator. Like
most electric equipment, generators are limited by their current-carrying capability. Near rated
voltage, this capability becomes an MVA limit for the armature of the generator rather than a MW
limitation, shown as the armature heating limit in the figure. Production of reactive power involves
increasing the magnetic field to raise the generators terminal voltage. Increasing the magnetic field
requires increasing the current in the rotating field winding. This too is current limited, resulting in the
field-heating limit shown in the figure. Absorption of reactive power is limited by the magnetic-flux
pattern in the stator, which results in excessive heating of the stator-end iron, the core-end heating
limit. The synchronizing torque is also reduced when absorbing large amounts of reactive power,
which can also limit generator capability to reduce the chance of losing synchronism with the system.
Usually, a synchronous generator is mandated to absorb or produce reactive power in a band bounded
by limits R1and R2, shown by dotted lines in Figure 6.6. However, outside this band, it is entitled for
opportunity costs if real power output is required to be reduced in order to produce more reactive
power. For example, as shown in the figure, if a generator is asked to supply reactive power equal to
B2, rather than B1, the operating point of generator changes from current point X to new point Y,
forcing reduction in real power output from D1 to D2. More on calculation of loss of opportunity cost is
provided in further sections.

Figure 6.6: Generator Capability Curve

Under the vertically integrated structure, only the capital and operating costs that could be associated
with the extra equipment (e.g., parts of the voltage regulator, exciter, stator, and rotor plus the
operating costs associated with field losses) required for voltage control were charged to the voltage
control function. However, in a deregulated environment, the opportunity costs associated with
reduced real power sales when excessive reactive power is required becomes an important component
of the total cost of providing voltage control from generators. In [23], it is mentioned that the costs
and prices for voltage support will be highly nonlinear with system load. At very high levels of system
load, the opportunity cost of voltage support will far exceed the embedded cost.

Synchronous Condensers

Synchronous machines that are designed exclusively to provide reactive support are called
synchronous condensers. Synchronous condensers have all of the response speed and controllability
advantages of generators without the need to construct the rest of the power plant. As compared to
the static alternatives of reactive power support, due to presence of moving parts and system
auxiliaries, they require more maintenance. They also consume real power equal to about 3% of the
machines reactive-power rating.

Capacitors and Inductors

Capacitors and inductors are passive devices that generate or absorb reactive power. They accomplish
this without significant real-power losses or operating costs. The output of capacitors and inductors is
proportional to the square of the voltage. Capacitor banks are composed of individual capacitors. The
individual capacitors are connected in series and parallel to obtain the desired capacitor-bank voltage
and capacity rating. The capacitor banks are often configured with several steps to provide a limited
amount of variable control. Inductors are designed to absorb a specific amount of reactive power at a
specific voltage. They can be switched on or off but offer no variable control.

Static VAR Compensators (SVCs)

An SVC combines conventional capacitors and inductors with fast switching capability so as to provide
a continuous range of control. The range can be designed to span from absorbing to generating
reactive power. Consequently, the controls can be designed to provide very fast and effective reactive
support and voltage control. Because SVCs use capacitors, they suffer from the degradation in
reactive capability as voltage drops.

Static Synchronous Compensators (STATCOMs)

The STATCOM is a solid-state shunt device that generates or absorbs reactive power and is one
member of a family of (FACTS) devices. The STATCOM shows similar performance as that of SVC,
when compared on the basis of response speed, control capabilities, and the use of power electronics.
However, the basic difference is that the STATCOM uses power electronics to synthesize the reactive
power output, without employing capacitors and inductors. The STATCOM ensures very fast and
efficient voltage control by virtue of its solid state nature. STATCOM capacity does not suffer as
seriously as SVCs and capacitors do from degraded voltage. STATCOMs are current limited - so their
MVAR capability responds linearly to voltage as opposed to the voltage-squared relationship of SVCs
and capacitors. This attribute greatly increases the usefulness of STATCOMs in preventing voltage
collapse.

Comparison between Various Sources of Reactive Power

The comparison can be based on three parameters: voltage support ability, costs and means of
procurement.

Voltage support ability

A limiting characteristic of capacitors and capacitor-based SVCs is that output drops dramatically when
voltage is low and support is needed most. STATCOMs provide more support under low-voltage
conditions than capacitors or SVCs because their capability drops linearly with voltage. The output of
rotating machinery (i.e., generators and synchronous condensers) generally rises with dropping
voltage. Generators and synchronous condensers generally have additional emergency capacity that
can be used for a limited time. Thus, Voltage-control characteristics favor the use of generators and
synchronous condensers.

Costs

Costs, on the other hand, favor capacitors. Generators have extremely high capital costs because they
are designed to produce real power, rather than reactive power. Even the incremental cost of reactive
support from generators is high, although it is difficult to unambiguously separate reactive power
costs from real power costs. Operating costs for generators are high as well because they can involve
real-power losses. Finally, because generators have other uses, they experience opportunity costs
when called on simultaneously to provide high levels of reactive and real power. Synchronous
condensers have the same characteristics as generators; but, because they are built solely to provide
reactive support, their capital costs are not as high and they incur no opportunity costs. SVCs and
STATCOMs are high cost devices as well, although their operating costs are lower than those for
synchronous condensers and generators.

Means of procurement

System operators can acquire reactive sources either through mandates or purchases. It might be
possible to create competitive markets for obtaining these services, provided the reactive supplies are
not geographically restricted. It is a general belief that the location limitations on reactive resources
are sufficiently demanding that competitive markets can not develop for this service. However, some
possible reactive power market designs are proposed in [23].Some system operators pay generators
their embedded costs for reactive resources. However, determining the embedded costs of generator
to provide reactive power support leads to ambiguity. This is so because; the same equipment is used
to provide both real and reactive power. Questions like what percentages, for example, of the exciter,
generator stator, generator rotor, turbine assembly, and step-up transformer should be assigned to
each function are not easy to answer. Table 6.1 shows the comparison of various types of reactive
power sources.

Ability Costs
Reactive
Speed of to
Power response support Operating Opportunity
Source
voltage
Synchronous
Fast Excellent High Yes
Generator

Synchronous
Fast Excellent High No
Condenser

Poor,
Slow,
capacitor Drops None No
Stepped
with V2

Poor,
SVC Fast Drops Moderate No
with V2

Fair,
STATCOM Fast Drops Moderate No
with V

Table 6.1: Comparison of various sources of reactive power support

Issues in Reactive Power Management

The research pertaining to reactive power management can be classified into two streams: market
design for reactive power and reactive power placement analysis. The possibility of establishing a
market mechanism for reactive power is an involved issue because of the basic difference in needs
and characteristics of reactive power when compared with real power. It has been mentioned earlier
that reactive power support should preferably a local phenomenon. Thus, it is now easy to appreciate
that the worth of 1 MVAR of reactive power is different at different locations in the system. Some of
these concerns about reactive power make market modeling for it different from that of real power.
These aspects are discussed at length with some alternatives of reactive power market designs in [5,
13, 16]. A great deal of research has been done on optimal allocation and sizing of reactive power
sources to improve the system voltage profile and reduce losses [6-12]. Reference [21] proposed an
integrated framework for optimal reactive power planning and its spot pricing, in which the selection
of VAR source sites is based only on the real power generation operation benefit-to-cost ratio for a
capacitor on a load bus. The approach neglects the effect of voltage improvement and system loss
reduction in the selection of VAR source sites. Three parallel methods can be used to determine the
potential sites for new VAR sources, namely: the cost-benefit analysis (CBA), the sensitivity method
(SM) and the voltage security margin method (VSMM) [21]. All three methods (CBA, SM and VSMM)
reflect the improvement of the system operation state after the VAR support service is provided. The
reader is referred to appropriate references for details of the above aspects.

BLACK START CAPABILITY SERVICE

A blackout is a rare contingency, but it nevertheless does occur. In order to reduce the economical
and social consequences, it is important to restore power as fast as possible. The system operator is
then bestowed with the responsibility of restoring the system to normal operating state as soon as
possible. However, restoration of the system after a major blackout is not an arbitrary process, but
needs a methodic sequential approach. Restarting of large thermal power plants requires major chunk
of electric power for its auxiliaries. On the other hand, the electric power resources like hydro plants,
diesel generators, etc., can be started without help from the grid. The system operator is required to
have enough of these resources at its disposal after blackout.During the restoration process, the
energization of long transmission lines and the capability of generators to support reactive power
creates major problem. Sometimes, under the deregulated environment, the restoration process may
involve private generators and multiple transmission companies. In such situations, the financial
compensation for these private entities adds a new dimension to the restoration process.The overall
coordination of network facilities owned by different entities and allocation of costs of various support
entities is hard to determine. Moreover, the worth associated with this type of support is not a
measurable quantity. Therefore, the system operator can make long term contracts so as to procure
black start capability. However, technical capabilities of contracted generators and their locations need
to be considered.

HOW TO OBTAIN ANCILLARY SERVICES?

The reason for existence of ancillary services is the security and reliability of the network. Smooth and
secured running of the system is of paramount importance, whether it be vertically integrated utility
or the restructured power system. Since one of the major aspects of deregulation is competition,
ancillary services can also be seen as an activity that can be subjected to competition. However,
whether the system security and reliability can go hand-in-hand with the competition, or rather,
should be subjected to competition or not is the moot question. Looking at the task achieved by each
type of ancillary services, one can easily make out that there are some services where competition can
be introduced, while for others, providing such services is mandatory for the participant. Thus, there
are two ways of obtaining the ancillary services: In the first case, provision of ancillary services is
made mandatory for the participant who wishes to take part in the energy market. The other way
could be to obtain the services on commercial basis. Here, the services can be obtained based on long
term contracts or daily competitive market.
It should be noted that introducing competition in the ancillary service market can not be directly
compared with the competition in the energy market. The system security is the most important goal,
and system operator should take all possible steps to achieve it. However, lack of competition should
not result in over payment by the system operator towards the ancillary services. Moreover, those
services which are inter-linked with the energy market (like spinning reserve), should not give chance
of market exploitation for few entities which are in a position to do so. Thus, competition in some
cases is necessary and fruitful. Both the ways of procuring ancillary services have their own
advantages and disadvantages. The choice of obtaining a service depends on the type of service and
practices prevailing in the system.

Mandatory provision of ancillary services

In this approach, before a participant is connected to the grid, it has to make sure that it is in a
position to provide the ancillary services mandated by the system operator. The system operator lays
down the rules to be followed by the participants. The rules for the connection to the grid can be:
The generator should be equipped with droop characteristics of 5%. This helps in frequency
regulation.
The generator should be able to operate in a power factor range of 0.85 lead to 0.9 lag. It
should be equipped with Automatic Voltage Regulator (AVR).

These types of compulsions act more or less like the rules existing in the vertically integrated utility.
This will ensure that enough resources will be available towards system security. This approach is a
simpler one; however, it does not lead to economic efficiency. Some of the problems associated with
this approach are as follows:
There is a chance that more than sufficient sources are likely to be developed, which is not
desired. For example, each and every generating unit need not take part in the frequency
regulation process.
The participants may think that they are denied the profits of the competitive market just
because they are forced to supply services at an additional cost.
The approach does not leave room for technological or commercial innovation.
Some units may be unable to provide some of the services. For example, nuclear power plants
can not be subjected to rapid changes in its output. Hence, same set of rules can not be
applied to all the participants.

Markets for ancillary services

The economic disadvantages and difficulty of practical implementation of compulsory services

necessitates introduction of competition in at least some of the ancillary services. The preferred form
of mechanism depends on the type of the ancillary service. Services like black start capability can be
procured on long term basis. These are the services in which the amount of service provided does not
change much with the time. Also, this amount does not depend on the activity of the spot market. The
system operator may run a separate market for regulation asking generators to submit their up and
down regulation bid. The reserve capabilities can be a mix of two approaches. The system operator
may make a long term contract for some part of the reserve requirement, while it can obtain
remaining reserve requirement through short term market mechanism.

Co-optimization of Energy and Reserve Services

In all the power markets of the early days, energy and reserve were traded in different markets.
These markets were cleared successively in a sequence determined by the speed of response of the
service. The market for reserve would clear first and then the energy market would be cleared. The
idea behind this was that the resources which were not cleared in one market would be offered in the
other market. Bids that were successful in one market would not be offered in the next market.It has
been mentioned in Chapter 3 that various products offered by a generator are strongly interlinked.
The natural choice of allocating various products is to go for a joint optimization of all products. Thus,
energy and reserve should be offered in joint markets and these markets should be cleared
simultaneously to minimize the overall cost of reserve and energy. If the energy and reserve are
obtained through separate markets instead of joint optimization, following situation may arise:
Partially loaded generators cannot sell as much energy as they might otherwise do.
 Out of merit order dispatch is done to meet the load.

The joint-optimization (also known as co-optimization) will take dispatch decision taking into
consideration the availability of resources and the overall cost. The joint optimization of energy and
reserve is done in multi-settlement nodal pricing markets like PJM.Suppose the energy and reserve
market is being co-optimized. As in the energy market, reserve providers offer to provide reserve. The
reserve offers do not reflect the cost of lost opportunity to sell the energy. Instead, they are related to
the expected cost of providing reserves, which might include some fixed administrative costs and
some variable operating costs associated with providing the reserve. Let us see what is meant by loss
of opportunity cost in the context of reserve provision with the help of Figures 6.7 and 6.8. Suppose
five generators are present in the system and they have provided the block bids.

Figure 6.7: energy-only dispatch

 Figure 6.8: Loss of opportunity cost

Let us work on standard simplifying assumptions and jump to results such that market clearing price
is with generator D as marginal generator. Now take the case of combined energy and reserve
market. Suppose RD is the system total reserve demand. Suppose, in addition to energy offer,
generating unit C offers to provide reserve for free, i.e., at 0 INR / MWh (this is a simplistic case we
have assumed). Then, generator C will have to de-load its scheduled output to PGc-RD . In order to
satisfy the power balance, generation D will have to increase its power output to P Gc+RD . The shaded
portion of Figure 6.7 represents loss of opportunity cost of generating unit C:(-C3).RD
Suppose there is a market that operates on a centralized basis, the generators bids to produce
electrical energy are equal to their marginal costs and that the market rules do not include separate
bids for the provision of reserve. To clear the market, the operator must determine the dispatch that
minimizes the cost of production while respecting operational constraints. For the sake of simplicity,
let us assume that the network has infinite capacity. Thus, the optimization problem can be
formulated as:

.........................................................................................................6.1
......................................................................................................6.2
.....................................................................................................6.3
.............................................................................................6.4
...................................................................................................................6.5
The Lagrangian function of the above formulation can be written as:

..................................................................6.6
............................................................................6.7
...............................................................................................6.8
By applying standard Kuhn-Tucker conditions:

..................................................................................................6.9
 .....................................................................................6.10
Thus, expressions for 1 and 2 can be given as:

............................................................................................................6.11
................................................................................................6.12

Thus, the Lagrange multiplier associated with the constraint on the production-demand balance gives
the marginal cost of producing electrical energy. Similarly, the multiplier associated with the reserve
requirement constraint gives the marginal cost of providing the reserve. Suppose, no limits are hit,
1 = 2 =3 = 4 =0 then, . This is equal criterion of conventional economic dispatch problem. All
generators work on the same marginal cost and marginal cost of providing reserve is zero. Suppose, a
particular generator i hits its maximum limit such that

.........................................................................................................6.13

then, and and expression for marginal energy costs and marginal reserve costs can be given as:

...........................................................................................................6.14
...................................................................................................6.15

From the last equation it becomes clear that the revenue collected by a generator is exactly equal to
opportunity cost of not selling the energy. Thus, even if a generator is asked to be available to provide
a reserve and thus hits its maximum limit including the energy generation, it does not incur monetary
loss because the system marginal cost for energy will be higher than that generators own marginal
cost of production and the difference will be paid to generator which can be assumed to be a loss of
opportunity cost. In a sense, the generator is indifferent whether it is asked to produce energy or be
available for reserve.

Thus, it is possible to clear energy and reserve markets simultaneously in a way that minimizes the
cost to consumers, meets the security requirements but also ensures fair treatment to all the
providers of energy and reserve services.

INTERNATIONAL COMPARISON
In this section, a brief account of technical and economic aspects of ancillary services of the markets
of England and Wales, Nordic Countries, California, New York and Australia is given. Aspects of two
important ancillary services, i.e., voltage control and frequency regulation are discussed.

England and Wales

The transmission grid of England and Wales is operated by National Grid Company (NGC), which is
also responsible for managing ancillary services. The costs associated with the provision of ancillary
services are recovered from the consumers through uplift in transmission payment mechanism.

Voltage Control

Provision of this service is required from all the generation units with a capacity over 30 MW. Voltage
must be kept within the range of Vnom+- 10% for 400, 275 and 132 kV networks and the power factor
(PF) between 0.85 capacitive and 0.95 inductive. The main costs associated with this service
correspond to capacity and operation. The ratio of capacity price to operation price is approximately
1:2. Consumers pay for the service through an uplift in their electricity tariff.

Frequency Control

The system operator (SO) is required to keep the frequency between 49.5 and 50.5 Hz. In the case of
a contingency, the frequency is allowed to drop under 49.5 Hz, but for not more than 1 minute. The
supply of primary frequency regulation is mandatory for all the generators with an installed capacity of
over 50MW, which must provide the service in continuous time. Generators are set for a droop of 3
5%. Large consumers can provide primary frequency regulation through load shedding, where a
response period of 10 s and a duration period of 20 s are required. The costs associated with the
service are generally traded through annual bilateral contracts. Consumers pay for this service
through an increase in their electricity payment which includes capacity, operation and compensation.
The service of secondary frequency regulation is considered as a commercial service and it is not
considered as a mandatory provision. Generators provide it using AGC. For those that provide the
service, a response time of 30 s and a duration period of 30 min are required. As in the previous case,
the costs associated with this service are difficult to identify, and they are traded based on annual
bilateral contracts and competitive auctions. The consumers pay through an increase in their
electricity tariff which includes capacity, operation and compensation.

Nordic Countries

Ancillary services are traded in a real-time market managed by the transmission system operators.
The costs associated with ancillary services are directly transferred to the users through the
transmission payment.

Voltage Control

The local provision of voltage control and reactive power is mandatory for all the system generators. A
response time of 5 s is required and the service must be supported for the period that is needed. The
costs associated with the service are low, since the provision is in charge of the hydraulic generators,
and the transaction mechanism corresponds to annual contracts at a fixed price. The suppliers receive
an operational payment, only for quantities that exceed the mandatory range and the consumers
payment is incorporated in the transmission tariff.

Frequency Control

The system frequency is required to be within the range of 49.5 and 50.5 Hz. The service of primary
frequency regulation must be provided within 30 s after being requested by the system operator and
must be supported for the period that is needed. The required quantity is locally calculated in every
country. This is an obligatory service for all the generation units, where they are required to keep a
droop between 2 and 5%. The transaction mechanism corresponds to annual contracts. The suppliers
get a payment for capacity and other for operation, and the consumers pay through the transmission
tariff. For example, Statnett, the system operator in Norway, determines a total amount to be paid
and each generator receives a payment according to its contribution to the service provision. The
service of secondary frequency regulation is only considered an ancillary service when it is provided in
the case of a contingency. For this service, a response time of 15 minutes and a minimal provision
period of 4 hours is required. The costs associated with secondary frequency regulation are minor,
since this provision is in charge of hydraulic generators. The provision is assured through competitive
offers. The consumers pay through the transmission tariff.

California, USA

After the power market crisis of 2001, the Californian ISO (CAISO) redesigned its electric energy
market. It then introduced the concept of available capacity (ACAP), whose objective is to allow the
ISO to verify in advance the availability of enough resources to satisfy the customer load as well as
reserves. In the new market structure, the energy market, the ancillary service market and the
congestion management market are jointly optimized.

Voltage Control

In the California system, the ISO procures reactive power support services on long-term contracts
from reliable must-run generating units. The actual short-term requirement is determined on a day-
ahead basis, after the real power market is settled and the energy demand and schedules are known.
Thereafter the ISO determines the location-wise amount of reactive power required based on system
power flow analysis. Daily voltage schedules are issued to contracted generators and the transmission
operators within the region. The generators are mandated to provide reactive power within the power
factor range of 0.90 lag to 0.95 lead. For reactive power absorption / generation beyond these limits,
the generators are financially compensated for, including, a payment if they are required to reduce
their real power output.

Frequency Control

Primary frequency regulation and secondary frequency regulation are not mandatory services. There
exists up and down service regulation. The costs associated with this service correspond to capital,
operation, fuel, and reduced efficiency costs, etc. For secondary frequency regulation, the service
must be available in 10 min and should be supported for at least 2 hours. The system operator
calculates the quantity required of the service according to a criteria that takes care of demand and
contingency. The quantity required generally comes close to 3% of the maximum demand of the
system.

New York, USA

In the New York control area, ancillary services are provided by the New York ISO (NYISO) or
procured by the transmission customers and suppliers themselves. The NYISO coordinates the
provision of all ancillary services and directly arranges for those services that are not self-supplied.
Depending on the nature of the services, either market-based or embedded cost-based prices are
used to price these services. In Table 6.2, the service provider and pricing method for each service are
given. Transmission customers and suppliers are permitted to self-supply operating reserve service,
regulation and frequency response service. They must bid the required resource into ancillary services
market. The NYISO selects the successful bidders to provide each service. The transmission customers
and suppliers with resources selected by the ISO use the revenues that they would otherwise have
received for providing these services, as an offset against charges they would otherwise need to pay
the NYISO for the service.

Service
Ancillary Service Pricing Method
Provider

Scheduling System Control Embedded Cost

NYISO
and Dispatch based

Embedded Cost
Voltage Support NYISO
based

Regulation and Frequenvy

NYISO or SS Market based
Response
 Energy Reserve NYISO Market based

Embedded Cost
Ooperating Reserve NYISO or SS
based

Table 2: NYISO Ancillary Service details

Voltage Control

Generating resources, which operate within their capability limits, are directed by NYISO to produce /
absorb reactive power to maintain voltages within their limits. The pricing method for the reactive
power support service is an embedded cost. The cost of reactive power support includes the following:

The total annual embedded cost for payment.

Any applicable lost opportunity cost to provide reactive power service.
Total of prior year payments to suppliers of reactive power service less the total of payments
received by the NYISO from transmission customers in the prior year for reactive power
service.

Loss of Opportunity Cost

If the NYISO dispatches or directs a generator to reduce its real power output in order to allow the
unit to produce or absorb more reactive power, the generator may receive a component of payment
accounting for the Lost Opportunity Cost (LOC). The concept of loss of opportunity cost was explained
with the help of Figure 6.6.

In NYISO, the method for calculating the LOC is based on the following factors:
Real time long-term based marginal price (LBMP).
Original real power dispatch and the new dispatch point.
Bid curve of generator supplying reactive power service.

Figure 6.9 describes the calculation of the lost opportunity cost for a generator which decreases its
real power output to provide more reactive power service. PRT is the Long-term Based Marginal Price
(LBMP), and f(P) is the bid curve of the generator supplying reactive power support. D1 and D2 are
the original and new dispatch points respectively while B1 and B2 are the corresponding bid prices at
D1 and D2. This sum is divided by the annual forecasted transmission usage for the year as projected
by the NYISO including the system load, exports and power wheeling. NYISO calculates this payment
hourly. Transmission customers engaged in power wheeling or exporting, pay the NYISO a charge,
which is equal to the value determined as mentioned above, multiplied by the wheeled energy at that
hour.
As the real power output is decreased, the generator receives lesser revenue from the sale of energy,
although, by way of this reduced generation, it saves some generating cost. The reduced income for
the generator (R) can be described by equation 6.16.

R = Revenue Loss Savings from reduced generating cost

......................................................................................................6.16

The first term in 6.16 denotes the revenue lost by the generator while backing down its real power
output from D1 to D2, and the second term denotes the corresponding reduction in generation cost.
Note that, R also equals the savings to the ISO. The saving of the generator (S) from reduced real
power output is described in equation 6.17.

.....................................................................................................6.17

The LOC of the generator equals to the difference between R and S.Thus, it is given as:

....................................................................................................6.18

Figure 6.9: Calculation of LOC

Frequency Control

This service is accomplished by committing on-line generators, predominately through the use of
automatic generation control, to follow moment-to-moment changes in load. Regulation service is bid
into the market by individual units that have AGC capability and that wish to participate in the
regulation market. Bid information includes regulation response rate (MW/min) and regulation
availability rate ($/MW). The NYISO selects regulation service in the day-ahead market from all
bidders.The bids submitted by suppliers are stacked from the lowest priced offer to the highest priced,
taking into account the operational requirements for up and down regulation. The NYISO selects the
bids starting with the lowest bid, and the last selected bid is set as market clearing price. All suppliers
selected in the market receive an availability payment calculated with the corresponding market
clearing price. The NYISO makes the following settlement with suppliers of regulation service:
An hourly availability payment for reserving capability to provide regulation service.
An energy payment based on the amount of regulation provided.

Australia

Initially, the ancillary services were traded through long term bilateral contracts between National
Electricity Market Management Company (NEMMCO) and the service suppliers. Since 2001, frequency
control ancillary services are traded in competitive spot markets.

Voltage Control

The energy code of Australia requires the power factor to be between 0.9 inductive and 0.93
capacitive. Optionally, generators can produce and absorb reactive power or operate the generation
unit as a synchronous condenser. The service provision is managed through annual bilateral contracts.

Frequency Control

In this service, it is expected that the systems frequency varies only in the range of 50 +- 0.1 Hz. The
system operator can request additional amounts of primary frequency regulation, up or down. A
response period from 6 to 60 s is required and the service should be supplied during 90 s. Generators
with AGC are in charge of the service provision which is managed through annual bilateral contracts.
The suppliers get payments for enabling and compensation. Secondary frequency regulation is
provided by generators, load shedding or fast connection generation units. This service is required in a
period of 5 min and the period by which the service should be supplied is not specified. The provision
of the service is managed through annual bilateral contracts. Payments are carried out for enabling
and compensation.

Module 7: Pricing of Transmission Network Usage and Loss

Allocation
INTRODUCTION

Transmission pricing and loss allocation are highly debated issues after the deregulation of power
industry. In the post deregulated era, the transmission provision gets a good deal of importance.
Strong transmission system forms the backbone of any successful deregulated power industry. As per
the planning policies developed in most of the countries, more emphasis was given on adding more
generation to the system rather than improving and strengthening the transmission network. But after
restructuring of the power industry, the issue of open access has compelled policy makers to re-think
their approach towards transmission planning. Open access demands sound transmission corridor
availability for a transaction to become viable. After deregulation, the transmission system is owned
and operated by a separate company that is popularly known as Transco . For well known reasons, the
transmission activity remains a monopoly rather than being a competitive activity. And since open
access demands a non-discriminatory access to the transmission system by any qualified entity in the
business, this monopoly entity has to be regulated by a higher governmental agency. In many
countries, the Transco are the dis-aggregated part of the then original vertically integrated utility that
existed in the region, prior to deregulation.

The original concept of reforms and restructuring was mainly aimed at gaining economic benefits to all
dispersed entities in the market. However, success or failure of a market depends on the design of its
market rules. The power sector reforms are full of uncertainties and due to its large scale, wrong or
inappropriate market rules may bring disaster instead of the economic gains at large. Transmission
pricing rules form one of the important parts of the market rules. As we all know, one of the basic
prerequisites of establishing a competitive market is the transmission open access. Successful
competition at the generation level calls for a successful, fair and non-discriminatory open access for
the transacting entities in the market. Pricing of transmission services plays an important role in
determining whether providing transmission open access and allied services is economically beneficial
to both the wheeling utility and wheeling customers.

Few years back, electricity transmission pricing was more of an academic interest, rather than
practical use. This is because generation, transmission and distribution were vertically integrated. The
vertically integrated utilities used to sell their power inside their territory, or exchange power with the
neighboring utilities. Hence, the need for having a formal mechanism for pricing of transmission did
not exist. The costs incurred by the vertically integrated utilities were recovered by embedding them
in the electricity price billed to the consumers. However, in recent times, as a primary step towards
reforms, generation and transmission businesses have been separated from each other in many
countries and hence, transmission prices are used to charge the transactions. One important fact
about the transmission pricing issue is that it is a technical issue rather than an engineering problem
[8]. To some extent it couples operational aspect of the power system with it, so long as provision of
correct economic signals is considered. Engineering analysis which deals mainly with determining the
feasibility and the cost of providing transmission services is only one of the many considerations in the
overall process of pricing transmission services.

Loss allocation is another issue, where no single solution can be ultimate solution. The loss allocation
problem is different from loss supply. Loss allocation is all about allocating costs of losses amongst
various participants. Loss allocation problem is a contentious issue because of the non-linearity
associated with power flows. We will see more about loss allocation methods towards the end of the
chapter.

What is Power Wheeling?

To define formally, wheeling is the transmission of electrical energy from a buyer to a seller, through
transmission or distribution lines owned by a third party. Call a selling utility as Utility S, the buying
utility, Utility B. Suppose they are non-contiguous and they are connected by several parallel paths
through different utilities in between. One of the utilities connecting them is Utility K. In the context of
the term wheeling, the obvious question to be asked is, what does it mean to say that Utility S is
wheeling to Utility B, via Utility K? The answer to this question is not straightforward where every
other entity in the business has its own perception and that is why the issue of wheeling and its
pricing becomes highly debatable. Two issues are highly debated upon [6].

Displacement or Movement?

It is a popular argument that power doesn't actually move from injecting point to load point. Rather,
displacement of power takes place by addition and withdrawal of some MW of power at injecting and
take-off points respectively. One analogy given by proponents of this concept is that of a lake that is
filled to its maximum capacity. Pouring water at one end of a lake makes an equivalent amount of flow
at the other end. However, the poured water does not travel from one end to the other. The electricity
pushed into a power pool would also behave in a similar fashion. Thus, it is argued that the
transmission users should pay a flat fee regardless of distance between two points.

The opponents of this view state that suppose x gallons of water is added at one end of the lake, it
may not travel all the way to the other end, but it causes incremental disturbance in every gallon in
the lake. Similarly, pushing some MWs at one point and withdrawing from the other would create an
equivalent total change in the system flow. Hence, the user should pay as if its power had been
moved over this much distance.

Distance Dependency: High or Low?

One of the debating point is how much transmission delivery price should depend on distance? This
question did not exist in case of a vertically integrated utility as all generation as well as loads were
looked after by a single utility. In this case, neither a load near generating station got advantage nor
the distant loads felt the brunt of their location.

But in a de-regulated era, the buyers have the choice to choose their generator as local or distant
producers. There is an argument that the customers should pay towards transmission usage on the
basis of the distance. However, the counter argument to this is that, why should a customer be
penalized for its geographical location, which is not in its hand? Thus, one aspect of this distance
parameter is the fairness. The other aspect is the generation of price signals

From the economists' viewpoint, low price as a function of distance creates a competitive market,
which is the motivating factor for de-regulation. On the other hand, high long distance transmission
charges would encourage construction of local generation so as to discourage long distance
transmission of power, leading to loss reduction. In other words, high dependency of prices on
distance would lead to engineering efficiency.

In spite of having different opinions on the debatable issues, one thing is agreed upon by everybody
and that is the charges for transmission system use should cover all the costs and provide a small,
regulated level of profit for the owners of the transmission facility. Thus, the real debate over pricing
focuses on how the cost of the system is allocated among its users? Apart from cost recovery, can the
pricing provide any other information? Based on this, some principles of transmission pricing have
been developed, which are discussed next.

PRINCIPLES OF TRANSMISSION PRICING

To operate the power system under the regime of transmission open access, a trade-off has to be
solved: Economic marketing of energy has to be given importance while at the same time; it should
be ensured that the whole system operates in a reliable and secure manner. The main purpose of any
transmission pricing scheme is not limited to recovery of the sunk costs involved in bringing up the
transmission infrastructure. The transmission pricing scheme should do much more than that. In line
with the above, following principles should be followed while designing the transmission pricing
schemes [3]:

1. The transmission prices should be devised so as to promote the efficiency of day-to-day

operation of bulk power market.
2. The transmission prices should signal locational advantages for investment in generation and
demand.
3. They should signal the need for investment in the transmission system.
4. The transmission prices should recover the costs of existing transmission assets.
5. Transmission pricing mechanism should be simple and transparent.
6. The mechanism should be politically implementable.

Out of these, the first three objectives are concerned with derivation of appropriate economic signals
to either utility or the consumer. However, the fifth objective states that the signals should not be so
complicated that one can not decipher the same and react to it. Fourth and sixth objectives are
associated with the allocation strategy of the pricing mechanism. Briefly speaking, the first objective
speaks about the short term efficiency, numbers 2-4 with long term efficiency and 5, 6 with
implementation.

There are different transmission pricing mechanisms prevailing in different parts of the world. They
differ on a lot of parameters like: whether they use incremental methods to price the transactions or
they go for rolled-in cost methods; whether generator pays the wheeling charge or the consumer pays
for it, or both pay a part of it in some proportion, etc. It is expected that while designing a
transmission pricing mechanism, following cost components for providing transmission service should
be taken into account [11]:;
1. Operating Cost: This includes the cost mainly due to generator rescheduling, maintaining
system voltage, reactive power support and line flow limits.
2. Opportunity Cost: It is the cost which a transmission company (Transco) has to forgo due to
operating constraints that are caused by the transmission transaction.
3. Reinforcement Cost: This cost is charged to only firm transactions and includes capital cost of
new facilities required to meet the transaction.
4. Existing System Cost: The investment cost of existing transmission facilities used by the
transmission transaction.

CLASSIFICATION OF TRANSMISSION PRICING METHODS

Almost all existing and proposed transmission pricing models are cost based. That means, they
allocate all or part of the existing and new transmission systems to wheeling customers. Based on
this, transmission pricing paradigms can be defined which convert the transmission costs into
transmission charges [10]. Three basic paradigms are:

Rolled-in (embedded) transmission pricing

Marginal transmission Pricing
Composite transmission pricing

We have seen in Chapter 3 that the power markets throughout the world are classified based on two
dispatch philosophies: centralized dispatch and decentralized dispatch . The decentralized dispatch
markets are the ones in which rolled-in paradigm of transmission pricing is commonly employed. On
the other hand, the centralized dispatch markets employ the marginal or the composite pricing
paradigm.

An alternative way of classifying transmission pricing schemes is based on when they are calculated,
i.e., ex-ante or ex-post . In the ex-ante schemes, the entities taking part into the power market
activities know the transmission prices a priori. While, in ex-post schemes, the transmission charges
are calculated only after the real time has elapsed and power flow snap-shot is available. These
schemes can further be categorized into transaction based and non-transaction based. The transaction
based schemes essentially should have a defined source point and a sink point (bilateral transaction).
On the other hand, non-transaction based schemes refer to the power exchange (PX) trades, where it
is not possible to identify source-sink pair. Figure 7.1 shows the broad categorization of various
transmission pricing schemes.
 Figure 7.1: Classification of transmission pricing schemes

In the above figure, the transmission pricing schemes are classified on the basis of whether they are
calculated ex-ante or ex-post. Generally, the ex-ante schemes are made up of pricing methods under
rolled-in paradigm. As mentioned earlier, the total costs to be recovered are known a-priori and then
they are transformed into transmission prices. The ex-post schemes, on the other hand, rely upon the
incremental or marginal pricing mechanism. Moreover, the incremental schemes lack the property of
recovering transmission sunk costs and hence rely upon schemes under the domain of rolled-in
paradigm to overcome this lacuna. This gives rise to the composite paradigm.

ROLLED-IN TRANSMISSION PRICING METHODS

In this paradigm, all the costs incurred during building the infrastructure and the future investment,
operating, maintenance costs are summed up (rolled-in) together and then are allocated to various
wheeling customers on various basis. The basic philosophy behind this paradigm of transmission
pricing paradigm is shown in Figure 7.2.
 Figure 7.2: Rolled-in Paradigm

Effectively, this boils down to directly or indirectly quantifying the extent of usage of the network by
each transaction. The diversity of underlying assumptions, methodologies, etc. lead to many choices
or versions of methods under this category. Some of the commonly practiced methods are as follows:

1. Postage Stamp Method (transaction / non-transaction)

2. Contract Path Method (transaction based)
3. Distance Based MW-Mile Method (transaction based)
4. Power Flow Based MW-Mile Method (transaction based)
5. Power flow tracing based on proportionate sharing principle (non-transaction)
6. Equivalent bilateral exchange (EBE) method (non-transaction)
7. Zbus based method (non-transaction)

There are some methods that allocate costs to individual bilateral transactions. These methods are
known as transaction based methods. On the other hand the rest of the methods allocate the total
costs to all the participants of the pool. These methods are called as non-transaction based methods.
All these methods will be explained one by one with an illustrative example in the following sub-
sections.

Postage Stamp Method

Postage stamp methodology is the simplest and easy to implement methodology of transmission
pricing. A postage stamp rate is a fixed charge per unit of power transmitted within a particular zone.
The rate does not take into account the distance involved in the wheeling. There are various versions
of postage stamp methodology. In some versions, both, generators and loads are charged for
transmission usage, while in others, only loads pay for the same. Some variants charge loads for their
peak value while in others, they are charged on the basis of average loads. A simpler version of
postage stamp mechanism is explained with the help of following illustration.
 Figure 7.3: Sample 5 bus system

Suppose that the rolled-in cost of a region INR1000/day and that there are bilateral transactions as
shown in Table 7.1.

From
Sr.No To Bus MW
Bus

1 1 4 40

2 2 2 15

3 2 3 35

4 1 5 50

5 2 3 10

Table 7.1: Bilateral Transactions

There are various ways of expressing the postage stamp rates. Normally it is given in INR/ MW/ day
for Indian system. Let us assume that the loads make the whole payment towards the transmission
charges. Then, the transmission charges paid by each load will be proportional to its MW. Hence, the
transmission price paid per day by each load will be as given in Table 7.2.

Charge per Day

Sr.No Load
(INR)

1 15 100

2 45 300

3 40 266.67

4 50 333.33
 Table 7.2: Transmission Charges by Postage Stamp Method

The postage stamp rates are based on average system costs and may have a variety of rate designs
based on energy charges, capacity charges, or both. Rates may include separate charges for peak and
off-peak periods, may vary by seasons and in some cases may be different for weekdays and
weekends.
Some of the advantages of Postage Stamp Method are as follows:

The method is simple and easy to implement.

It is transparent and is easily understood by all.
There is no mathematical rigor involved.
Recovers sunk cost of transmission system.
Being very simple and straightforward, it is easy to get political backing for it to be
implemented.

Disadvantages of the Postage Stamp Method can be quoted as follows:

Pancaking: In case a transaction takes place such that the power is transmitted through
multiple intermittent utilities or zones, pancaking of access charges takes place.
No economic signal: With regard to the principles discussed in the earlier sections, postage
stamp allocation does not create an economic signal associated with the effect of a particular
transaction.
No extent of use of network: Postage stamp allocation does not take into consideration the
extent of use of the network by a particular transaction. The transmission charges paid by two
loads, out of which, one is very near to a generator, while the other is miles apart, is the
same. It is obvious that transmission network use by the other load is more than the first.

Incremental Postage Stamp Methodology

To reduce the effect of pancaking, an alternate version of postage stamp method, called Incremental
Postage Stamp Method is sometimes employed. An incremental postage stamp rate could be applied
to a zone which is much smaller than a region. This avoids pan-caking in the case of inter-regional
transactions. If an incremental postage stamp rate is assigned to a zone of Xkm x Xkm , then the
charges for distance greater than Xm would become sensitive to distance.

In Figure 7.4, suppose a bilateral contract exists between entities A and B, where A is situated in area
P and B is situated in area Q. Suppose stamp of area P is INR 90/MW and that of area Q is INR
180/MW. Then the transaction pays a charge of INR 270/MW towards wheeling. Now consider the
second figure. Both areas, P and Q are divided into sub-areas and each horizontal block of 3 sub-areas
is assigned an equal postage stamp. That means, each sub-area in area P has a postage stamp of INR
10/MW and each sub-area in Q has postage stamp of INR 20/MW. Now, the transaction between A
and B has to pay a charge of INR 30/MW towards wheeling, instead of INR 270/MW in the earlier case.
 Figure 7.4: Postage Stamp and Incremental Postage Stamp Methods

Contract Path Methodology

This method is based on charging the transacting entities between two points, based on a pre-defined
path. To define formally, contract path is the shortest route formed by a series of transmission lines
which can carry the contract power between the take-off point and injection point. In the earlier days,
when the wheeling contracts were rare, the contracts were used to be written between the utility and
the contracting parties. Hence, the word contract in the name. Even though contracting parties know
that the power will split into multiple parallel paths, they compute the prices for a single path.

For example, assume that power is being wheeled from the state of Orissa to the state of GU as
shown in Figure 7.5. The contract path would be as shown in the left hand side figure. However, there
are bound to be multiple parallel paths when power flows from Orissa to GU, as shown in the right
hand side of the figure. To get the feel of how charges are calculated under this scheme, assume the
contract is for 100 MW, with the contract path as shown in the figure. Then, if the capacity of Orissa-
CH corridor is 200 MW, this transaction will be charged 50 % of the cost of this line. Similarly, if
capacity of CH-MH corridor is 100 MW, the whole cost of the line would be attributed to this
transaction.

Figure 7.5: Contract Path Method

This method is viable, rather justified, in a system where the network is not so cramped and hence, it
is easy to determine the length of a contracted path to a great extent in such cases. Suppose the
transactions of Table 7.1 are to be charged as per the contract path, then Table 7.3 shows the
contract paths and associated transmission charges for each of the transactions are shown in Table
7.4.

Transmission
Charges
Line Capicity (MW)
(INR/MW/Time
Duration)

1-2 50 1.0

2-3 70 1.5

3-4 40 0.75

3-5 10 0.5

1-5 100 1.0

4-5 50 1.0

Table 7.3: Transmission Charge Corridor

Transaction Transaction
Sr. Contract Transmission
No
ChargeCalculation
Between Amount
Path Charges

1 1-4 40 1-2-3-4 40 x (1+1.5+0.75) 130

2 2-2 15 - - -

3 2-3 35 2-3 35 x (1.5) 52.5

4 1-5 50 1-5 50 x (1) 50

5 1-3 10 1-2-3 10 x (1+1.5) 25

Table 7.4: Contract Path Charges

The prices for transmission lines will be in INR/MW/Duration time. Some of the advantages of this
methodology are as follows:

This method is simple to implement, though not as simple as postage stamp rate method.
irectly or indirectly, the method takes into account the distance involved in wheeling.
Avoids pancaking to a large extent.

Drawbacks of this methodology are as follows:

The contract path between the points of take-off and injection is decided a priori without doing any
simulation. Hence, the method fails to provide correct economic signal. Power flows can not be
restricted to a particular path if parallel paths are available. For example, in Figure 7.3, transaction
between generator on bus 1 and load on bus 4 has three parallel paths, i.e., path 1-5-4, path 1-2-3-4
and path 1-2-3-5-4. Hence, it is hard to decide which path the power would flow. Now suppose that
three parallel paths in this particular case are owned by two different transmission owners. For
example, lines 1-2, 2-3, 3-4, 3-5 are owned by utility 1 and lines 1-5, 5-4 are owned by utility 2.
Then, in that case, Genco of Bus 1 and Load at Bus 4 make a contract with, say, utility 1 for contract
path, utility 2 is in loss. This is because; its network asset is being used, but is not being paid back,
just because a contract is not done.

MW-Mile Methodology

This method overcomes the limitations of the contract path method. This method bases the cost on a
computed set of parallel paths for a particular transaction. The method employs power flow simulation
to determine the flow of a transacted power in various lines. There are two versions of the MW-Mile
methodology. The first version which is a non-power flow method is an approximate method. It does
not make use of any power flow simulation.

Distance Based MW-Mile Methodology

Distance based MW-Mile method evaluates the usage of each user according to the product of the
quantity of the transacted power and the geographical distance between the source and sink. This is
quite a rough method. In practice, due to the effect of meshed network, there is no fixed relationship
between the geographical distance and the actual costs.

This is a simple, easy to calculate and easy to apply method. It can be called as a version of the
contract path method. However, this method does not take into consideration the effect of actual
power flow and transmission users do not face their actual costs. This does not lead to efficient
operation of power systems.

Power Flow Based MW-Mile Methodology

Power flow based MW-Mile method takes into account both the quantity of transacted power and the
electrical distance between source and sink and allocates the total costs in proportion to the MW-Mile
of transactions. There are various versions of power flow based MW-Mile methods. The original MW-
Mile method suggested by [8] states that:

Given a transaction with the actual points and the variation of generation and load specified, MW-Mile
methodology calculates the maximum transaction related power flow on every transmission line using
a DC power flow.

For transaction t, according to MW-Mile methodology, real power flows on all network lines are
calculated using the DC power flow algorithm. The magnitude of MW flow on every line is then
multiplied by its length L1 and a predetermined weighting factor reflecting the cost per unit capacity of
the line, W1and summed over all network lines. This leads to:

......................................................................................................................(7.1)

This process is repeated for every transaction by considering only the generations and loads
associated with that transaction. The share of the total transmission network capacity cost, TC,
allocated to transaction t can be calculated according to the following formula:
 ......................................................................................................................(7.2)

The main characteristic of MW-Mile methodology is that we need to determine the usage or influence
of each transaction on each branch. Based on this, there are various versions of MW-Mile approach.
The concept of power flow based MW-Mile method is explained with the help of illustrative example of
Figure 7.6. The figure shows a 3 bus system.

Figure 7.6: Illustrative example for MW-Mile methodology

Details about the network elements are given in Table 7.5.

Reactance
Sr.
Element Ll Wl Pmax
No
p.u

1 1-2 X12=0.2 200 0.25 100

2 3-2 X32=0.25 100 0.5 50

3 1-3 X13=0.4 100 0.5 50

Table 7.5: Details of network elements

Total transmission network capacity cost is calculated as follows:

................................................................................................................7.3

Thus, TC comes out to be INR 10,000. Assume that base case power flows are zero. Suppose two
bilateral transactions are taking place as shown in the figure:

1. T1= 65 MW between generator at bus 1 and load at bus 2

 2. T1= 65 MW between generator at bus 1 and load at bus 2

Line Line Line

MWMilet TCt
1 -2 3-2 1-3
Flow
(MW)
50 15 15
T1=65
4000 6400
MW MWMile
2500 750 750
t,l

Flow
10
(MW)
25 10
T1=65
2250 3600
MW MWMile 5001-
1250 100
t,l 3

Since this methodology allocates transmission charges based on maximum usage of a transmission
line by a transaction, it emulates the actual planning process for system reinforcements which is
based on local considerations rather than coincident peak condition for the overall system.
The advantages of this method can be stated as follows:

It is insensitive to the order of wheeling transactions. This is because every transaction is

treated separately by considering only those generators and loads that are associated with
that transaction. Hence, there will be no dispute about the order in which the transactions
should be considered.
It gives a correct signal to both short distance and long distance entities, unlike in postage
stamp case.
The method is intuitively logical and conceptually straightforward.

Some of the drawbacks of the above methodology are:

Since the method uses DC approximation of the power system, it leads to inaccuracy in
calculating the extent of use of the network by a particular transaction. This is because, the
real power system is modeled by a set of non-linear equations.
No merit is attributed to the transactions which give rise to counter flows, thereby reducing
loading of the system.

Power Flow Tracing Based Network Fixed Cost Allocation Method

The power flow tracing methods notionally quantify the usage of the network elements by various
generators and loads. By making use of this information, the network fixed costs can be allotted to
various entities. In general, by notional decomposition of transmission line flows and losses, the
tracing algorithms provide following information:

Contribution of kth generator in meeting jth load.

 Loss incurred while transferring kth generators power to jth load.
Decomposition of power flows on a line into its constituent generators and loads.
Losses supplied by various generators.
Losses due to various loads.

Currently, two well accepted power tracing approaches are available: one is proposed by Bialek [1]
that is based on simultaneous equations approach and the other is graph theoretic approach proposed
by Kirschen et. al. [19] and Wu et. al. [18]. These methods are based on proportionate sharing
principle. The pre-requisite for tracing is a valid power flow solution. Thus, after the power flow
solution is obtained using KCL and KVL, the above tracing algorithms invoke the proportionate sharing
rule to solve the problem.

Proportionate Sharing Principle

The proportional sharing principle is explained with the help of Figure 7.7.

Figure 7.7: Proportionate sharing principle

Here, four lines are connected to node i, two with inflows and two with outflows. The total power flow
through the node is Pi=40+60= 100 MW
of which 40% is supplied by line j-i and 60% by line k-i. As electricity is indistinguishable and each of
the outflows down the line from node i is dependent only on the voltage gradient and impedance of
the line, it may be assumed that each MW leaving the node contains the same proportion of inflows as
total nodal flow Pi. Hence, the 70 MW out flowing in line i-m consist o f 70 x 40/100 supplied by line j-i
and 70 x 60/100=42 MW supplied by line k-i. Similarly the 30MW out flowing in line i-l 30 x
40/100=12 MW consists of supplied by line j-i and 30 x 60/100 = 18 MW supplied by line k-i.

The proportional sharing principle basically amounts to assuming that the network node is a perfect
mixer of incoming flows so that it is impossible to tell which particular inflowing electron goes into
which particular outgoing line. In [17], it is shown that though proportionate sharing principle
simplifies the problem, it is not an inherent characteristic of tracing problem and the solution space
can be characterized by treating this problem as an optimization problem. In the next section we will
see the details of graph theoretic approach. The graph theoretic approach is quite intuitive and easy to
appreciate. The details of the simultaneous equations approach are not provided here and the reader
is encouraged to see reference [1] for more details.

The Graph Theoretic Approach

The graph theoretic algorithm has two versions: upstream-looking algorithm and downstream-looking
algorithm. The upstream-looking algorithm will allocate the transmission usage/supplement charge to
individual loads and conversely, the downstream-looking algorithm will allocate the transmission
usage/supplement charge to individual generators. The algorithm is constructed on a matrix
formulation and therefore enables the use of linear algebra tools to investigate numerical properties of
the algorithm.

The downstream tracing of graph theoretic approach starts at a pure source. Pure source node is
defined as a node which has no transmission line inflows associated with it. It corresponds to a node
with the highest node angle ( ). Downstream algorithm is a tops-down approach where nodes are
deleted symbolically in a sequential manner. It is assumed that there are no circular flows in the
system. In the downstream algorithm, nodes are eliminated in the descending order of the node
angles. A node elimination operation involves the following:
1. deletion of the node and its associated components like lines and generators.
2. insertion of tagged fictitious generation at receiving end to maintain flow equivalence with the
rest of the network. For this purpose, the conventional algorithms have used strategies based
on the proportionate sharing rule.

The step by step algorithm to carry out downstream tracing is given as follows:

Step1: Start with a pure source node. A pure source is defined as a node on which only real power
outflows exist and there are no inflows of real power.
Step2: Insert tagged fictitious generation at receiving end m as follows:

.
.................................................................................................................(7.4)
This can be explained with the help of Figure 7.8. In this figure, if virtual generation of 40 on bus l is
to be inserted as tagged fictitious generator on bus m, then its value would be
40/43 x 43/100x 40=16.

Figure 7.8: Downstream tracing concept

Step3: Let and be sending and receiving end powers in MW on line lm. Let be the
fictitious generation in MW of generator k, on node l, which is either physically present there or has

come out of deletion of the earlier node. Now, the loss in the line lm, MW can be allocated
to generator k as follows:
 ................................................................................................................... 7.5

where is the number of generators present (physically or virtually) on node l. In a similar

manner, the line flow can be attributed to generator k.

Step4:After each deletion of a pure source, there is at least one pure source node left in the system.
Delete this pure source node and repeat step 2.

Step5: Repeat step 2, 3 and 4 till all the nodes left in the system are pure sinks. A pure sink node is
defined as a node which has no transmission line outflows associated with it. At the end of step 4, one

of the results obtained is , which is nothing but the net power received by load i at its bus, sent
from generator k.
The upstream algorithm works exactly in the same manner, but in a reverse order. It starts from a
pure sink and terminates at a pure source.

The method is applied to a sample system with the power flow solution as shown in Figure 7.3. As
explained above, the nodes are deleted one by one, starting with a pure source, which is node 1. The
nodes are deleted in the following sequence: 1, 2, 3, 5. Node 4 is the only pure sink node in the
system. Generator on bus 1 is designated as Generator A and generator on bus 2 is designated as
Generator B. When node 1 is deleted, it has two lines connected to it on which the outflow exists: Line
1-2 and Line 1-5. It is obvious that only Generator 1 can contribute to the power flowing on these
lines. So, a fictitious generation of 29.018 MW which corresponds to Generator A, is attached to Bus 2.
Similarly, a fictitious generation of 70.982 MW is attached to Bus 5. In the next step, Bus 2 becomes a
pure source node. It is deleted and generations corresponding to Generator 1 (29.018 MW) and
Generator 2 (35 MW, after supplying local load of 15 MW) are carried on proportionality basis over the
line connected to Bus 2, i.e. line 2-3. For example, we know from load flow solution that the flow on
line 2-3 is 64.018 MW. So, amount of Generator 1s power flowing on line 2-3 will be:

Similarly, amount of Generator 2s share in power flow on line 2-3 is:

The above calculated fictitious generations corresponding to each generator is attached to Bus 3.
Then, Bus No. 3 is deleted. It has two outflows: Line 3-4 and Line 3-5. So, the amount of generator
1s power flow on line 3-4 will be:

Similarly, amount of Generator 2s share in power flow on line 3-4 is:

The load on Bus 3 is also shared in a similar proportionality based manner. Share of Generator 1 in
load on Bus 3 is given by:

Similarly, Generator 2s share in load on Bus 3 is given as:

In a similar manner, contributions of generators on rest of the loads and lines can be found out. Table
7.7 furnishes the results in which decomposition of transmission line power flows into various
generators contribution, for the five bus system.

Line Gen1 Gen2

1-2 29.018 0
1-5 70.982 0
2-3 29.018 35
3-4 8.348 10.07
3-5 0.272 0.328
5-4 21.45 0.098

Table 7.7: Decomposition of transmission line flows

Table 7.8 gives the results of downstream algorithm for generator contribution in loads on the five bus
system. Once line decomposition is obtained as shown in table 7.7, the costs of lines can be attributed
in linear proportion to the load and generators participation in the transmission line flows.

L1 L2 L3 L4
Gen
0 20.3975 29.8 49.801
1

Gen
15 24.6024 10.2 0.1982
2

Table 7.8: Contribution of Generators in Loads

Equivalent Bilateral Exchange (EBE) Method

The power flow tracing methods employ proportionate sharing rule on each bus to decompose the
transmission line in-flows into out-flows. Thus, they invoke an additional rule of proportionality at each
bus to find out contributions of generators and loads in transmission line flows and losses. The results
are worked out on the power flow solution and are topology based. So, all generators do not
contribute to all loads and a generator or a load may contribute to flows of only some lines and not all.
The equivalent bilateral exchange (EBE) model [14] works on similar principles but the proportionality
is assumed in global terms. It works on the assumption that each generator supplies fraction of each
load in proportion to its system contribution compared to total generation. Each demand is supplied by
a fraction of each generator uniformly divided among all generators. Similarly, each generator supplies
a fraction of each demand uniformly divided among all demands. The method works on a DC power
solution that neglects losses. The method is explained as follows:

Step1: Obtain a DC power flow solution

Step2: Let be the fraction of generation that supplies the load . Then,

s ...................................................................................................................(7.6)
Where PD is the total system demand,

..................................................................................................................(7.7)
Thus,

......................................................................................................................(7.8)

......................................................................................................................(7.9)
With this decomposition, the effect of the equivalent bilateral exchange GDij, on the power flow in line
lm can be determined uniquely.

Step3: Solve a DC load flow consisting of a single power injection of GDij at bus i and the load at bus
j. The power flow on line lm due to this equivalent bilateral exchange can be given as:

.................................................................................................................................(7.10)
Step4: The costs of the line lm can be attributed to this equivalent bilateral exchange in linear
proportion to the usage

The solution thus obtained is independent of the choice of slack-bus.

Z-bus Cost Allocation Methods

This method has been proposed by Conejo et. al. [12]. This method uses the contributions of the
nodal currents to line power flows to apportion the costs of the transmission network. Once a load flow
solution is available, the proposed method determines how line flows depend on nodal currents. This
result is then used to allocate network costs to generators and demands.

Consider Figure 7.9. Let Sj,k be the complex power flow computed at bus j and flowing through the
line connecting bus j to bus k. Let us define the power flow direction from bus j to bus k as positive.
Then,

.....................................................................................................................................(7.11)
Using the Z-bus matrix, the voltage at node j is given by
 .................................................................................................................................(7.12)

The current, is obtained as

....................................................................................................................(7.13)
Substituting (7.12) in (7.13),

..................................................................................................(7.14)
Re-arranging the above equation,

.....................................................................................................(7.15)
The first term in the above equation depends on the network parameters and can be written as

....................................................................................................................(7.16)
Where,

Thus, can be termed as a measure of electrical distance between bus i and line jk. The active
power through any line can be decomposed and allocated to nodal current injections as follows:

................................................................................................................(7.17)

And the transmission line cost allocation can be done based on the factor that shows contribution of
ith node in real power flow over line jk.

Figure 7.9: equivalent of line jk

There is one point of contention though. The factor is defined such that Pjk > 0. However, the
factor can also be defined as expression for shows that , with opposite convention for power
flow direction. The

expression for shows that

Thus, definition of positive direction of flow will decide the contribution of node i in power flow of a
line.

The rolled-in transmission pricing paradigm is considered to be economically inefficient. This is

because it ignores the transmission resources scarcity. For example, a new transmission transaction
that causes major new transmission reinforcements due to capacity bottlenecks is likely to be
inefficient even if the cost of energy is very low. Rolled-in transmission pricing paradigm may not
indicate this inefficiency since the cost of new reinforcements is spread among all energy customers.

In general, the methods in this paradigm are even though simple to understand and implement, fail to
satisfy the other principles of transmission pricing.

MARGINAL TRANSMISSION PRICING PARADIGM

According to this paradigm, only the new transmission costs caused by the new transmission
customers will be considered for evaluating transmission charges for these customers. The existing
system costs will remain the responsibility of utilities' present customers. The basic philosophy lying
behind this paradigm is shown in Figure 7.10.

Figure 7.10: Incremental Paradigm

In contrast to the philosophy of rolled-in methods, the embedded costs (sunk costs) are not taken into
account, but the additional transmission cost a transaction causes is attributed to the transaction
itself. Under this scheme, marginal cost of energy is calculated, which includes loss and network
constraint components. Depending on the time-frame under consideration, the marginal costs can be
classified as short-run and long-run.

As we have seen in Chapter 2, for short-run considerations, the transmission capacity is considered to
be fixed, while for long-run pricing schemes, it is assumed that new capacities can be built. Thus,
long-run marginal cost includes the reinforcement and expansion cost as well as the operating cost.
Short-run marginal costs only reflect the operating costs of the existing facilities

Short-run Marginal Cost Pricing (SRMC)

The foundation of this methodology is based on the theory of spot pricing [9]. Under the marginal
pricing schemes, the general idea is to model an electricity market with its various economical and
technical specifications such as generators' cost functions, demand elasticity, generation limits, power
flow limits, etc. Then, this system is optimized with the objective of maximizing the social welfare. An
important outcome of this optimization is the nodal price at each load, which is popularly known as
Locational Marginal Price. It reflects the temporal and spatial variation of the energy price relating to
the energy demand [9]. Because of losses and system security (line limits), a kWh of energy has
different values at different busses of the network. Since wheeling is analogous to buying energy at
one set of buses and selling it at another set, these spatial price differences determine the cost of
wheeling. The optimal spot price at bus i defined as in [2] is given as:
{cost of additional demand at ith bus} =
{cost of additional demand at swing bus} x{1+ incremental losses caused by i} + {transmission
constrained terms summed over lines}
We have seen the details of nodal spot price (or LMP) calculation in Chapter 5. Recalling the simplistic
expression for spot price,

.........................................................................................................(7.18)
Where,
nj......Shadow price of line j, nj=0 if line flow limit is not reached
..... System lambda
i...... Optimal spot price at bus i
di....Demand at bus i
L.....Transmission Losses
Zj.....Line flow in line j

With no transmission constraints and losses neglected, there will be only one system lambda, i.e. one
system price for the whole network. As this methodology comprehends that the electricity not only has
to be generated, but also has to be delivered to a particular load, taking into account transmission
constraints and electrical losses, the difference in nodal prices then gives the marginal network
revenues (NR), which is nothing but the SRMC for that particular transaction. To define formally, the
marginal operating cost per MW of transacted power can be estimated as the difference in the optimal
cost of power at all points of delivery and receipt of that transaction. The marginal operating cost is
then multiplied by the magnitude of the transacted power to yield the SRMC for the transmission
transaction. It can be given by the following equation:

..................................................................................................................(7.19)

where, i is the bus i marginal cost, Pi,t, the injected power at bus i due to transaction t and Bt the
set of transmission buses involved in the transaction t. This is explained with an illustrative example of
Figure 7.11.
 Figure 7.11: Incremental Paradigm

The is indicate the short run nodal prices of bus i. If bilateral transactions of Table 7.1 are assumed
to take place, then each transaction pays a charge as shown in Table 7.9.

Difference Transmission
Tranction
Sr.No of Nodal Charges
Between
prices (INR)
1 1-4 0.57 22.8

2 2-2 0 0

3 2-3 0.13 5.85

4 1-5 0.44 22

Table 7.9: Wheeling charges as per marginal pricing scheme

With the marginal cost based pricing method, it is possible that the transmission fixed costs
(embedded or sunk costs) may not be recovered fully. This is because, no constraint in the network
results in zero network revenues (if lossless model is considered). This will not be acceptable to the
transmission owner company. As stated in [5], the spot pricing recovers the total incurred network
costs only in the case where the installed system capacity is optimal. Hence, to recover the sunk
costs, a top-up of the charges is required to be done. This leads to the composite paradigm of pricing
which is explained later.

Long-run Marginal Cost Pricing (LRMC)

In the calculation of SRMC, it was assumed that the transmission capacity is fixed. For calculation of
LRMC, this assumption is removed. To define formally, the LRMC are the costs of increasing the
production by one unit, allowing changes in the overall system capacity, i.e., reinforcing the system.
For the optimal capacity of the network, LRMC and SRMC are equal. In this pricing methodology, the
marginal operating and reinforcement costs of the power system are used to determine the prices for
a transmission transaction. Marginal operating cost was defined in the earlier section. The marginal
reinforcement cost is calculated as follows [10]:

Over a long time horizon of several years, all transmission expansion projects are identified and their
costs are taken into account. This cost is then divided over the total power magnitude of all new
planned transactions to calculate the marginal reinforcement cost.

In other words, the LRMC of a transmission transaction is the sensitivity of the network capacity cost
to the quantity of the transmitted power, that is, the least network reinforcement costs needed by per
unit increment of power transmission. Accurate calculation of LRMC is very difficult and some
assumptions and simplifications are done. These are as follows:

the capacity of transmission lines can be increased continuously.

there are no new right of ways
the peak demand condition is considered.
all the lines are of the same type.
the costs for lines are linear functions of branch capacities.

Under these assumptions, the LRMC becomes the sensitivity of the system MW-Mile to the quantity of
the transmitted power [1].

COMPOSITE PRICING PARADIGM

In this paradigm, price of a transmission service is determined based on the sum of the total of
embedded and incremental costs of providing the service. The embedded cost of a transmission
transaction is part of the existing system cost that is allocated to that transaction. Figure 7.12
illustrates the basic idea behind this paradigm. In this paradigm, the embedded charges of a
transmission transaction are evaluated based on one of the allocation methodologies described in
section 7.4 and the incremental cost is calculated using one of the methodologies described in section
7.5.
 Figure 7.12: Composite paradigm

The recovery of fixed costs is done through a complementary charge defined as follows:

....................................................................................................................(7.20)
where,NRl is the marginal annual revenue for line , is the complementary charge for line . It has been
proved in [5] that the complementary charge is necessary to implement the marginal pricing scheme.
For allocation of complementary charge, three alternative schemes have been proposed in [7]. These
are as follows:

1. Adjustment of Marginal Prices: Nodal prices are modified according to some criterion (an
additive term, a multiplicative term, Ramsey prices, etc.), in order to match the marginal
network revenues and the total network cost.
2. Extent of Use Allocation: The complementary charge is allocated among the various entities
depending on their extent of use of the network.
3. Benefit Allocation: The complementary charge is assigned among the agents depending on
the economic benefit that each of them obtains from each network facility

Due to a huge gap between marginal network revenues and network sunk costs, the option of
adjustment of marginal prices is very seldom considered. One of the methods commonly used in some
of the South American markets the marginal participation method, is explained next.

Marginal Participation Method;

In this method, incremental usage of the network is considered and the fixed costs are distributed
using the marginal participation factors. The usage by a node is defined as the change in power flow of
a line or a corridor after incrementing 1 MW load (or generation) on that bus. This increment is
distributed amongst all other buses in pro-rata manner while calculating the power flow. Thus, the
power flow variation in each line incurred by each node is obtained. This exercise can be repeated for
multiple scenarios. The usage index for each network user is established as follows:
 ..................................................................................................................(7,21)
Where,Fce is the power flow of the corridor c at scenario e; Fce k is the power flow in corridor c in
scenario e after 1 MW increase is carried out on node k. Pke is the power consumed or generated at
node k. Ukc is usage factor of node k for corridor c. The marginal participation factor of a node k over
corridor c is given as:

..................................................................................................................(7,22)
The costs to be recovered for corridor c are then attributed to individual nodes in proportion to the
participation factors calculated above. The marginal participation method is employed in Chilean and
Argentinean systems.

COMPARISON BETWEEN DIFFERENT PARADIGMS

The pricing scheme of a perfectly competitive market is not applicable to transmission markets
because they exhibit the characteristics of natural monopoly. Hence, three different paradigms do
exist to resolve the issue of transmission pricing.The rolled-in paradigm has a clear cut advantage so
long as aspects like cost recovery and practicability are considered. Postage stamp allocation
methodology from this paradigm is one of the most popular methods of cost allocation. However,
rolled-in transmission pricing paradigm is considered to be economically inefficient since it ignores
transmission resource scarcity. For example, a new transmission transaction that causes major new
transmission reinforcements due to capacity bottlenecks is likely to be inefficient even if the cost of
energy is low. Rolled-in transmission pricing paradigm may not indicate this inefficiency since the cost
of new reinforcements is spread among all energy customers. The methods fail to set incentives for
either the efficient use of the system and for further expansion.
The marginal pricing schemes promote economic efficiency by setting incentives for the momentary
use of existing transmission facilities and for the reinforcement of the network. Despite its
advantages, the marginal pricing schemes are likely to fail in recovering the total (embedded)
transmission costs. Moreover, as the transmission revenues increase, the further is the system from
an optimal state. In that case, the ISO or the Transco may also have an incentive to withhold the
transmission capacity. One more de-merit attributed to this paradigm a few years back was that
calculation of incremental prices was a formidable task, as a good deal of computational rigor is
involved in calculating nodal prices.

Incremental
NO Principle
Rolled-in Incremental
Day - to -day
1
operation efficiency
Signal for
2 investment
generation
in
Signal for
3
investment in
 transmission

Cost recovery of
4
existing assests
Simplicity and
5
transparency
6 practicability

Table 7.10: Comparison of transmission pricing paradigms with respect to principles

But with the state-of-the-art development in computing facilities, this has become an issue of less
concern. In fact, some of the markets in USA need to calculate nodal prices through security
constrained economic dispatch for every 5 minutes. Table 7.10 shows the comparison of both the
transmission pricing paradigms with respect to the six principles mentioned in section 7.2.

DEBATED ISSUES IN TRANSMISSION PRICING

There are various issues associated with transmission cost allocation, the answers to which are system
specific. Some common concerns among these are: generator pays the wheeling charge or the
consumer, or both pay a part of it in some proportion; which system simulation state is employed for
cost allocation; how the cost allocation of unused capacities is done; how to price for counter flows,
etc. Diversity of answers to these questions leads to different versions of transmission pricing
schemes. It should be noted that most of the schemes providing solutions to these set of concerns
belong to a rolled-in or composite class of transmission pricing schemes. A review of transmission
fixed cost allocation methods which address some of the above issues is provided in [4]. A brief
account of the issues and solution choices to them is provided in this section.

1. Basis of system state: In this category, transmission cost allocation schemes are classified
on the basis of various system states that are considered. Further classification in this
category is as follows:
System peak state: In this, the costs are allocated among users according to their
transmission usage or benefit at the time of system peak flow [21, 22].
Weighted multi-state: In this method, the transmission usage and benefit of the user
at more than one system states are studied [23].
User peak state: In this method, the costs are allocated in proportion to the maximum
power flow of each user [8].
Maximum line flow: In these methods, the cost is allocated to various entities
whenever the flow on a line is at its maximum [24].
Game theory based: This method is based on the fact that the sum of maximum flow
caused by all users is larger than the system peak. Hence, to deal with the matter of
conflict, nucleolus scheme in the cooperative game theory is employed in the
allocation [25].
2. Cost Allocation of unused capacities:Any practical network is built with redundant
capacities to account for the network security. The unused transmission capacity is defined as
the difference of facility capacity and the actual flow on that capacity. The original MW-Mile
methodology [8] ensures the full recovery of all the embedded costs and assumes, inherently,
that all transmission users have to pay both for the actual capacity use and for the unused
transmission capacity. The objections to this approach are as follows [26]: First, this pricing
rule does not encourage more efficient use of transmission systems since, whatever the
 capacity utilized, the total cost of the transmission line is recovered. Second, some users may
find the cost allocation procedure as unfair when they have to share the cost of an expensive
transmission facility for which only a small portion of the facility capacity has been utilized
[27]. Third, the approach is not equitable in the presence of multi-buses transactions with
netting effect. In this category, the costs of unused capacities are allocated either on postage
stamp basis [5, 27] or cost of unused capacity of each branch [26]. Then, there are some
methods based on the reliability considerations. In [28], the costs of unused capacities of each
branch are allocated in proportion to the reliability benefit of the transaction.
3. Pricing of Counter Flows:Counter flows are those flows which have different direction than
the net flow on the line. In a sense, the counter flows tend to relieve the loading of a line.
Hence, the debatable question is whether to charge the transactions causing counter flows or
to credit them. Based on this, three categories of proposals are formed: credit the counter
flows; counter flows pay as per absolute value of flows; and counter flows pay zero.
4. Who pays, generators or loads:Decision about allocating total costs between generators
and loads is the policy decision taken by regulatory authorities. From the viewpoint of
promoting the long run investment of generators and loads, the costs should be allocated to
both generators and loads according to their usage of the network. In [27], it is mentioned
that the generators should be allocated more costs than the loads because their location is
more sensitive than the loads. Different viewpoint on this issue states that to minimize the
distortion of the demand charge to the economic signal provided by SRMC, according to the
concept of Ramsey pricing, more costs should be allocated to loads. In the league of European
nations (ETSO), the generators share in transmission price ranges from 0% to 50%.
5. The nodal pricing or the zonal pricing:The non-transaction based pricing schemes usually
charge an entity based on the MW injection (or drawal) into (or from) the grid. If this rate per
MW is same everywhere, the scheme takes the form of postage stamp method. However, if
the charges have spatial variation, the scheme is typically termed as Point-of-Connection
(POC) transmission tariff. The basic principle of POC is that payment at one point, i.e., the
point of connection gives access to the entire network. However, the charges vary by location
creating price signals. If the charges are applied to each and every point in the system, the
commercial settlements become too complex. To overcome this problem, the points can be
aggregated to form zones. Then, each zone will have its own POC tariff. Forming zones can
sometimes be an involved task. At some places, the charges are defined as per the
demarcation formed by the ownership of the network.

LOSS ALLOCATION
Loss allocation in the restructured era pertains to allocating the costs associated with losses to
individual users of the network. Loss allocation does not affect generation levels or power flows. It is
about distribution of revenues and payments amongst suppliers and consumers. Thus, loss allocation
should not be confused with loss supply. Loss supply is a mechanism by which the system losses are
generated when these are not accounted for during the original dispatch. The loss supply service
changes the network variables. On the other hand, loss allocation does not directly affect the network
variables. The losses are known from the system state and these are divided using some criteria
amongst the participants. Various loss allocation schemes have emerged from the desire to establish
fair criteria for loss allocation. The issue of loss allocation is contentious because there is a hint of
arbitrariness while choosing a rule or criteria. Let us see why.

Why Loss Allocation is Contentious

Losses are non-linear functions of line flows.If linearization techniques are used to allocate the flow of
a given line to generators and loads, the cross terms associated with quadratic functions do not allow
assigning losses to generators and consumers in a unique manner.Consider a single transmission line

where two transactions create power flows and respectively. The approximate power loss
on the line lm will be given by the following expression:

...................................................................................7.23

Where, R is the resistance of the line. A simple and most intuitive way of splitting the above losses can
be given as follows:

.....................................................................................7.24

This is also known as incremental loss allocation method. So, one way of allocating the losses is such
that in addition to its own quadratic contribution to the total losses, each transaction is allocated

exactly one half of the cross term, irrespective of its size.If flows on line lm due to two

transactions are comparable, i.e. then loss allocation can be considered to be equitable.
However, this rule becomes objectionable when the flows arising due to two transactions vary
significantly. Four typical situations showing combinations of two transactions are shown in Figure
7.13. When flow on the line is due to a single transaction, the total loss is equal to square of the
power flow; which is attributed solely to the transaction causing the power flow (cases a and b). Thus,
for 10 MW flow transaction, loss is 1 and that for 1 MW flow is 0.01 MW. However, while in case c, we
can see that loss shared by 10 MW and 1 MW flows are 1.1 and 0.11, respectively. Thus, change in
the losses for larger transaction is 10% while that for smaller transaction is 1000%. The transaction
with smaller flows will certainly not buy this.
Now consider case d. If we assume that the transactions creating counter-flows are credited (i.e., paid
for loss reduction), the higher level transaction may not be satisfied by the way the lower transaction
is getting benefited.

Figure 7.13: Various cases for loss allocation

Thus, assigning equal value associated with cross term will lead to discrepancy in loss allocation.
Further, if we change the rule for assigning cross term losses, the solution will change leading to
arbitrariness. So, the generic formulation for decomposition of cross-term losses can be given as:

.....................................................................................................7.25
which can be written as:

. .............................................................................................................7.26
Based on the above equations alternative schemes have been proposed to decompose the cross-term.
These are:

1-Proportional Allocation:

2-Quadratic Allocation:

3-Geometric Allocation:

Figure 7.14 shows plot of versus the ratio for different alternatives of allocation
discussed above.

Figure 7.14: Plot of ii Plm1 /Plm2

The incremental method, as discussed earlier has i,irrespective of value of the ratio . The

geometric allocation shows a linear change in with the ratio .The incremental allocation
bears an important property called aggregate invariance. It means that even if a load splits virtually
into smaller loads, the losses allotted to it remains the same. Thus, a load can not get advantage by
splitting into smaller fractions.Now the question is how the flows on lines are related with the nodal
injections? The essence of loss allocation by incremental methods is to expand the system losses
through a first order sensitivity with respect to the nodal generation and load injections. We will see
more about this under the section called incremental loss allocation methods.

Classification of Loss Allocation Methods

The loss allocation methods developed so far can be classified into three broad categories as follows:
1. Pro-rata loss allocation methods
2. Incremental loss allocation methods
3. Proportional sharing loss allocation methods

We will see some details of these methods one by one. It should be remembered that these loss
allocation methods are ex-post mechanisms and have nothing to do with the loss supply decisions in
the scheduling process.

Pro-rata Methods

In these methods, the total loss is first divided between two groups: generator losses and load losses.
The percentage losses in each category are pre-defined. The losses are then allocated to each
generator or load in proportion to its generation or consumption, respectively. Let,P L represent the
total system losses. Let G be the fraction of total losses allotted to generators and D be the fraction
of total losses allotted to loads, such that,

......................................................................................................(7.27)

...............................................................................................................(7.28)

Thus, losses allotted to individual generators and loads are given as:

..................................................................................................(7.29)

................................................................................................(7.30)
Where, we define KG and KD as generation and load loss allocation factors, respectively.

Incremental Methods
These methods calculate the incremental transmission loss (ITL) coefficients. The ITL is defined for a
bus and quantifies the change in total losses produced by an incremental change in the power injected
on that bus. The ITLs are obtained from a converged power flow solution.

..........................................................................................................(7.31)

Thus, straightforward loss allocation to generators i and loads j is:

.............................................................................................................(7.32)

.........................................................................................................(7.33)

It is easy to appreciate that due to non-linearity associated with power flows, the following situation
arises:

....................................................................................................(7.34)

So, we need to allocate the normalized losses such that sum of allotted losses equals total system
losses. Let normalized incremental transmission loss coefficient be defined as

.......................................................................................................(7.35)

Where

............................................................................................(7.36)

Thus,

.........................................................................................(7.37)
Hence, the loss allotted to individual generators and loads will be:

.........................................................................................................(7.38)
 ......................................................................................................(7.39)
It is worthwhile to note that the losses allotted to individual generators and loads may come out to be
negative. Another matter of debate about this method is that the results are dependent on the choice
of slack bus.

Power flow tracing based loss allocation

We have seen principles of tracing based on proportionate sharing principle in section 7.3. Equation
7.5 shows us how to allocate losses to individual generators. Thus, topology dependent loss allocation
is done using proportionate tracing methods. The loss allocation under this scheme is also called as
average loss allocation.

Other Methods

There are several other approaches for loss allocation which do not fit into any of the above three
classes [12, 29, 30]. Here, Z-Bus method developed in [12] is explained. The Z-bus loss allocation
method is based on expressing total system losses in simple manner related directly to the equations
describing a solved load flow condition. If all generators and loads are represented as current
injections into the system, total losses can be expressed according to:

.................................................................................................(7.40)

..............................................................................(7.41)

In [12], it is shown that in a network that can be represented by a symmetrical impedance matrix, the
second component in the above equation becomes zero. Thus, loss allotted to bus i is

...................................................................................................(7.42)

Comparison between various methods

It is easy to infer that the losses associated with a bus are dependent on the current injection as well
as relative location of that bus in the system. In [13], qualitative comparison of three typical methods
of loss allocation is provided: pro-rata methods, incremental transmission loss methods, proportionate
tracing methods. Table 7.11 provides this comparison along with properties of Z-Bus loss allocation.

NO CHARACTERSTICS PR ITL PT Z-BUS

1 Slack bus dependency No Yes No No

2 Negative losses No Yes No Yes

3 Network topology dependency No Yes Yes Yes

 4 Dependency on MW quality Yes Yes Yes Yes

5 Linearity requirenment Yes No Yes No

6 Ease of understanding 1 3 2 4

1=Maximum, 4= Minimum
PR=Pro-rata,ITL=Incremental,PT=Proportionate

Table 7.11: Comparison of loss allocation methods

Pro-rata methods are simple to understand, but they do not consider the network. Thus, loads far-
away from the generators are benefited as compared to those nearer ones. The greatest lacuna of ITL
based methods is that the ITL coefficients are slack-bus dependent. In addition to this, the ITL
coefficients can be negative or positive and hence, the losses allotted can be negative or positive. The
tracing based loss allocation is not slack-bus dependent and considers the network flow conditions.
However, linear proportion is not the only way to solve the problem and any rule (for example squared
proportion) can be employed to allocate losses. In [27] it is shown that if instead of linear proportion,
squared proportion is used, the decomposition becomes aggregate invariant. In other words, a single
entity gets benefited by artificially decomposing its load into multiple loads of smaller value. Thus,
linear proportion is considered as a simple and aggregate invariant way of decomposing the flows and
the losses.

The advantage associated with Z-Bus methods is that there are least assumptions and hence the
method is technically justifiable. On the other hand, as in ITL based method, it is difficult to stomach
the idea of negative losses when it comes to practical implementation.The comparison of the schemes
brings out an important fact that no single loss allocation method can be termed as an exact scheme.
This is because the loss allocation methods are given system specific considerations and what is most
appropriate under one condition may prove a shortcoming in the other system. For example, most of
the developing countries adopt some version of pro-rata methods for the sake of simplicity and easy
adoption. Well established and well behaved power markets may see this as a hurdle in creating a
competitive environment. The undermining point is, there is no ultimate generic solution as such and
choice of the methodology solely depends on the prevailing system conditions and practices.The loss
allocation schemes are directly coupled to fairness issue. And that is what precludes one to establish
the ultimate scheme, as the term fairness has subjective interpretation.

REFERENCES

1. J Bialek. Topological generation and load distribution factors for supplement charge
allocation in transmission open access. IEEE Trans. Power Syst., 12(3).

2. hhM. C. Caramanis, R. E. Bohn, and F. C. Schweppe. The costs of wheeling and optimal
wheeling rates. IEEE Trans. Power Syst., PWRS-1:63{73, 1996.

3. ichard Green. Electricity transmission pricing: an international comparison. Utilities

Policy, 6(3):177{184, 1997.

4. Z Jing, X Duan, F Wen, Y Ni, and F F Wu. Review of transmission fixed costs allocation
methods. In Proc. IEEE PES General Meet, pages 2585{2592, Toronto, 2003.

5. I. J. Perez-Arriaga, F. J. Rubio, J. F. Puerta, and J. Arceluz J. Marin. Marginal pricing of

transmission services: an analysis of cost recovery. IEEE Trans. Power Syst., 10(1).
6. L. Philipson and H. L. Willis. Understanding Electric Utilities and De-regulation. Marcel
Dekker, Inc., 1999.

7. J. Rubio-Oderiz and I. J. Perez-Arriaga. Marginal pricing of transmission services: a

comparative analysis of network cost allocation methods. IEEE Trans. Power Syst., 15(1)

8. D Shirmohammadi, P R Bribik, E T K Law, and J H Malinowski R E O'Donnell. Evaluation of

transmission network capacity use forwheeling transactions. IEEE Trans. Power Syst.,
PWRS-4:1405{1413, 1989.

9. Fred C. Schweppe, Michael C. Caramanis, Richard D. Tabors, and Roger E. Bohn. Spot
Pricing of Electricity. Kluwer Academic Publisher, 1988.

10. D. Shirmohammadi, V. Filho, B. Gorenstin, and M.V.P. Pereira. Some fundamental

technical concepts about cost based transmission pricing. IEEE Trans. Power Syst.,
11(2):1002{1008, 1996.

11. D Shirmohammadi, C Rajagopalan, E R Alward, and C L Thomas. Cost of transmission

transactions: An introduction. IEEE Trans. Power Syst., 6(3):1006{1016, 1991.

12. Conejo, A. J.; Contreras, J.; Lima, D. A.; Padilha-Feltrin, A.;. Zbus transmission network
cost allocation. IEEE Trans. Power Syst., 22(1), Feb. 2007.

13. Conejo, A.J.; Arroyo, J.M.; Alguacil, N.; Guijarro, A.L.. Transmission loss allocation: a
comparison of different practical algorithms. IEEE Trans. Power Syst., 17(3), Aug. 2002.

14. Galiana, F.D.; Conejo, A.J.; Gil, H.A., Transmission network cost allocation based on
equivalent bilateral exchanges. IEEE Trans. Power Syst., 18(4), Nov. 2003.

15. Jiuping Pan; Teklu, Y.; Rahman, S.; Jun, K., Review of usage-based transmission cost
allocation methods under open access. IEEE Trans. Power Syst., 15(4), Nov. 2000.

16.Gomez Exposito, A.; Riquelme Santos, J.M.; Gonzalez Garcia, T.; Ruiz Velasco, E.A., Fair
allocation of transmission power losses. IEEE Trans. Power Syst., 15(1), Feb. 2000.

17. Abhyankar, A.R.; Soman, S.A.; Khaparde, S.A.;, Optimization approach to real power
tracing: an application to transmission fixed cost allocation. IEEE Trans. Power Syst., 21(3),
Aug. 2006.

18. Wu, F.F.; Yixin Ni; Ping Wei, Power transfer allocation for open access using graph
theory-fundamentals and applications in systems without loopflow. IEEE Trans. Power
Syst., 15(3), Aug. 2000.

19. D. S. Kirschen, R. N. Allan, and G. Strbac, Contributions of individual generators to loads

and flows, IEEE Trans. Power Syst., 12(1), Feb. 1997.

20. J. W. M. Lima, Allocation of transmission fixed rates: An overview, IEEE Trans. Power
Syst., 11(3), Aug. 1996.

21. C. W. Yu, "Long-run marginal cost based pricing of interconnected system wheeling.",
Electric Power Systems Research., 50(3), Aug. 1999.

22. R.D.Tabors. "Transmission system management and pricing: new paradigm and
international comparisons," IEEE Trans. Power Systet, 9(1), Feb. 1994.

23. H. Rudnick. M. Soto and R. Palm. "Use of system approaches for transmission open
access pricing." Electrical Power and Energy Systems, 21(2) 1999.

24. D. Kirschen, R.Allan and s. Ahmed, Allocating transmission system usage on the basis of
traceable contributions of generators and loads to flows, IEEE Trans. Power Syst., 13(2),
May 1998

25. Y. Tsukamoto and I. Iyoda, Allocation of fixed transmission cost to wheeling

transactions by cooperative game theory, IEEE Trans. Power Syst., 11(2), May 1996

26. A. Zobian and M. D. Ilic, Unbundling of transmission and ancillary services II Cost based
pricing framework, IEEE Trans. Power Syst., 12(2), May 1997.

27. J. W. Bilalek and P.Kattuman, Real and reactive power tracking: Proof of concept and
feasibility study, Electric Power Research Institute, Tech. Report, TR112416, Feb.1999.

28. J. W. M. Lima, M. V. F. Pereira, and J. L. R. Pereira, An integrated framework for cost

allocation in a multi-owned transmission system,, IEEE Trans. Power Syst., 10(2), May
1995.

29. R. J. Camfield and A. G. Schuster, Pricing transmission services efficiently, The

Electricity Journal, 13(9), Nov 2000

30. C. W. Yu and A. K. David, Pricing transmission services in the context of industry

deregulation, IEEE Trans. Power Syst., 12(1), Feb 1997.

31. F. F. Wu, P. Varaiya, Coordinated multilateral trades for electric power networks:
Theory and implementation, Univ. California Energy Inst., Tech. Rep. PWP-03, 1995.

32. G. Gross, S. Tao, A physical-flow-based approach to allocating transmission losses in a

transaction framework, IEEE Trans. Power Syst., Vol 15, No. 2, May 2000.