Strategic Bidding and Generation Scheduling in Electricity Spot-Markets

Kankar Bhattacharya
Department of Electric Power Engineering
Chalmers University of Technology
S-412 96 Gothenburg, SWEDEN
Tel: +46-31-772 1651, Fax: +46-31-772 1633, E.Mail: [email protected]

Abstract: This paper presents a short-term generation scheduling In this market structure, the independent power
model applicable to independent power producers participating in producers (IPPs) have to deal with several complex issues
electricity spot-markets. The model combines unit commitment, arising from uncertainties in spot market prices, and technical
generation scheduling and spot-market buy-sell decisions in a constraints which need to be considered while scheduling
common profit-maximizing framework. A scheme is proposed generation and trading for the next day. In addition to finding
thereafter to determine the optimum bid prices for each hour by
maximizing the probability of a bid being cleared and of accruing dispatch and unit commitment decisions while maximizing its
positive returns from the market. The IEEE Reliability Test System profit, their scheduling models should include trading decisions
is modified and adopted to represent an example IPP. Six distinct like spot-market buy and sell. Subsequently, the problem of
bidding strategies dependent on IPP’s marginal cost and expected finding appropriate bidding strategies for the trading decisions
spot market price emerge from the studies. emerges. A desirable bidding strategy should consider the
Keywords: Power generation dispatch, Deregulation, Bidding complex inter-play between technical aspects of unit operation
strategies,Power systems like ramping, limits, minimum run-time, etc., the economic
interests of an IPP as well as the uncertainties associated with
1. INTRODUCTION spot-market trading.
The bilateral contract dominated markets, like those found in Here, we will confine our discussions to the Nordic
the Nordic countries and in California, have a distinct feature type market shucture where generation scheduling is
that generation scheduling activities are decentralized and are decentralized and is the responsibility of an IPP. We will
c a n i d out by power companies independently to suit their own attempt to develop suitable bidding strategies for an IPP
objectives and requirements. This is contrary to the British incorporating its short-term operational issues vis-&vis its
Power Pool, which operates as a centralized entity in participation in spot-market.
controlling the schedule and dispatch of generators. In addition Among some of the works in this area, energy
to the bilateral contracts, spot markets have also been brokerage system formulations were presented in [ 1,2] where
established in the bilateral contract dominated markets, for the cost quotations by each utility were determined from an LP
example NordPool in the Nordic where energy trading is formulation and the interchanges were set up to maximize the
carried out based on bids submitted by participating parties. As profits to each utility. Subsequently bidding for power
of now, about 25% of the energy consumption of Sweden is exchanges have been developed, in which buyers and sellers
traded in the NordPool spot market, and the volume is growing. participate in auctions (even double or multiple auctions) by
In NordPool, the market is settled daily at noon for delivery of inter-acting through an Independent Contract Administrator
power commencing from the midnight up to the next 24 hours. [3,4]. In [3] a double auction game is used in combination with
The participants submit their buy and sell bids on an hourly classical optimization and incorporating operational constraints
basis, which are then aggregated to form the total supply and to determine the energy transactions while in [4] a genetic
demand curves. The clearing price and quantity is determined algorithm is used to determine the bidding strategies. In [5] a
from the intersecting point of these two curves. All bidders, probability density function is used to characterize the
whose bids are cleared, get the uniform system price competitor’s bidding range and the optimal bid is that, which
irrespective of their bids. This is called indiscriminate auction maximizes the lower bound on expected value of savings. In
and is a better way to check for market imperfections, like [6] an analytical formulation of the problem faced by a bidder
guessing of other parties’ bids. However, market imperfections in competitive power pool of England and Wales is reported.
may still exist. The optimal bidding strategy developed therein is for the case
of perfect competition, taking into account uncertainty in load
forecast and actions of competitors. Bid solicitation and
selection methods in spot markets and the various biddable
components in energy and capacity is discussed in [7].
Paper accepted for presentationat the International
Conference on Electric Utility Deregulationand 1.1 Problem Formulation and Approach
Restructuringand Power Technologies2000, We assume that the IPP has a mix of customers- a part of its
City University, London, 4-7 April 2000. load is under direct bilateral contracts and their prices are
fixed aprion. Secondly. i t has an unrestricted option to
0-7803-5902-Xl00l%10.0002000 IEEE.
participate in spot-market as a buyer and/or a seller. It also
108
maintains certain reserve generation available at all hours to Apart from the above, there are other constraints on unit limits,
account for contingencies, load variation or spot-market price ramp rates, etc. which remain the same as in case of a
fluctuations. In this framework the IPP has to devise its conventional UC. Those are listed here:
operating plan for the next day and take decisions on the Ramp Rate Constraint on T h e m 1 Units
following- Minimum Up, Down-time Constraint on Thermal Units
a. The Unit Commitment (VC) schedule. Generation Limit
b. The trading level @uy/sell schedule) in spot market. Coordination Constraints
c. Based on the trading levels what should be the bid prices. Hydro Energy Availability over 24 hours: A simplistic
d. The operating strategy to choose. hydro scheduling constraint ensuring optimal allocation of
Its objectives could be many- but most importantly it the total hydro energy available over the day, is used here.
would seek to maximize profits. Other constraints like maximum and minimum water
The following assumptions can be made concerning the IPP discharge rate, hydraulic continuity equation, maximum
and the deregulated market environment. and minimum water storage limits, etc. can also be
IPP’s generators have linear cost curves and consequently, modeled depending on data availability and
constant marginal costs computational capabilities.
The market is not perfectly competitive. Other participants Must-Run Units
are sensitive to IPP’s bidding- based on their information, Initial States of Generating Units
forecast and estimates. This means, the JPP’s bids have an
impact on spot-market price (pk) to a certain extent. 3. STRATEGIC BIDDING
The IPP has a good short-term estimate of spot market 3.1 Marginal Cost (MC)
price, (pk) in order to formulate its strategies. The UC model in Section-2 provides the IPP with optimum
A three-step approach as explained below is used to trading levels i.e. buy and sell quantities for each hour and also
determine the IPP’s operating plan for the next day and the corresponding scheduled total generation. The next step is
formulate its appropriate spot-market bids: to determine appropriate bid prices for the trades. To formulate
Step-1: Solve the generalized price-based UC (Section-2) to bid prices the IPP should have an idea of its hourly MC to meet
obtain the next day’s UC, hydro schedule and hourly buy I the scheduled total generation. This is found from the dual of
sell transaction decisions. the demand-supply constraint of an ELD program with demand
Step-2: Fix the hydro schedule and UC decisions. Determine being replaced by scheduled total generation.
the IPP’s hourly marginal cost (MC) to meet the scheduled
buy I sell transactions as obtained in step-1. An economic 3.2 Strategy Options for Bids
load dispatch (ELD) program can be used. The bidding strategy to be adopted by the IPP will depend on
Step-3: Determine the optimal bid prices for the transactions its MC for an hour vis-&-visk.For example, if its MC is higher
using the scheme proposed in Section-3. than k, a sell bid above MC, can ensure positive returns. On
the other hand, for a high probability of acceptance of the sell
2. GENERALIZED PRICE BASED UC bid, the IPP needs to bid below p+.. If a buy bid has to be placed
Objective: Maximize the estimated net profit over a day. under the same conditions, bids below MC will provide positive
Profit = Revenue from [spot market sell + bilateral power sell] returns while bids above p+. will have high probability of
- Payment for [spot market buy + unit operating costs + acceptance. Also, the IPPs bids could affect the system price in
start-up COSIS + shut down COS~S] certain cases and it is likely that then they would affect the bids
This can be mathematicallyexpressed as (1). of other parties.
1 Under these circumstances, if the IPPs decision is to
sell b units of energy at an hour, how should it bid? One way is
to simultaneously consider the probability of acceptance of the
A nomenclature is given in the Appendix. bid, returns expected from the market and how the bid can
affect pk.
Constraints: Let’s assume that Pk is centered around the estimated
__
Demand-Supply - Balance: The IPP must meet its bilateral value, k i.e., Pk is a normally distributed random variable with
contracts and maintain certain committed reserve generation probability distribution function f(pk), mean at and standard
at each hour. It has the option to generate, buy or sell in the deviation Q (4). The cumulative distribution function of f(p3,
spot-market. F(pd is given by (5) and (6) where @(a)is the Laplace function
or probability integral with properties @(O)=O, @(-+-@(a),
cP(-)=l. The values of @(a)is obtained from standard Laplace
function tables [8].

109
 Equn.( 12) assumes that the market imperfections come into
play when BP is beyond the range (pk 0).Then, p is adjusted
linearly with the difference between BP and ( p + ~or ) (p-o).
Some standard BP with their corresponding PA and
P+R is given in Tables-1 and 2 for the case of sell bids and buy
bids respectively. It is to be noted that P i s corresponding to the
chosen BPs are standards for a normal distribution as described
by (4). For each case, P+Ris also given for three situations
depending on MC and p.
F(pd denotes the probability that Pk is less than or
equal to a bid price, BP, i.e., P(p@Pk). Thus, the probability PA p+R

that a sell bid, with price BPSell, being cleared in the spot- MC<u I MC=u I MC>u
Say, = p -o Say,=p+o
market is when Pk 2 BPSellk and given by (7). similarly, for a p-30 -3/42 0.99864 Eqn.8) Eqn.(9) Eqn49)
buy bid, with price BPBuy, the probability that it will be g-20 -a42 0.9769 Eqn49) Eq n49) Eq nd9)
cleared in the spot market is when P k S BPBuyk,given by (8). p-a -1142 0.8423 Eqn.(9) Eq n49) Eqn49)
1
PA(BPSellk) = P(BPSellk 5 p k ) =- 1 @(ak,BpSe]])] (7)
2[- 1 1 1 I
PA(BPBUyk) = P(BPBuyk 2 fi)
p+20 1 2/42 0.0231 1.0 1.0 1 .o
ti+3cs I 3/42 I 0.00136 1 1.0 I 1.0 1 10
1
=1-p(fi 2
5 BPBuYk)=-[1+ @(ak,BpBuy)] (8)

In addition to the bid being cleared, another concern for the IPP
is whether the bid will derive positive returns (+R) from the
market. The returns will also depend on the MC vis-%vis the pk
For example, sell bids yield +R when Pk > MC while buy bids
when P k < MC. Thus, the probability that a sell bid price
(BPSell) yields +R (P+R)is given by (9), while the probability
that a buy bid price (BPBW) yields +R is given by (10).
Simultaneous probability (PGAJN) that a bid is cleared and yields
+R is given by (1 1) and this is of real interest to the IPP.
1
P+R(BPSdlk)=P(pk > M C k ) = -2[ l - @ ( a ~ ~ , k ) ] (9) Using PA and corresponding a's from Tables-1 and
2, a 3rd-order polynomial curve fit can be used to find general
P+R(BPBUyk)=P(fi <MCk)=l-P(pk > M C k ) = L [I + Q)(aMC,k)](10) functional forms for PAfor both sell and buy bids, as given by
2
(13) and (14) respectively.
PGAIN(B~~)=PA(B~,).P+R(MC~) (1 1)

It is apparent from (1 1) that to increase P G for~its sell bids,

the IPP should either bid below Pk to ensure a high PA or bid
above MC to ensure a high P+R. For its buy bids, it should bid
above p k for a high PA or bid below MC to affect a high P+R.
~ ~ = 0.0504 a3 - 0.4584a + 0.5
P A - s ~(a)

P*(BPSell) =O.O504( BPSell

P A _ B(a)
UJi-

~ ~ = -0.0504 a3+ 0.4584a + 0.5

r
-' - 0.458,( BPSell
cfi
-')+ 0.5 (13)

However, adopting any such strategy is likely to affect the bids

of other parties, inducing them to bid lower or higher than their PA (BPBuy ) = -0.0504
expected bid, based on their estimate and knowledge of the
IPPs MC and likely bidding strategy. This will affect P k by way
of a reduction or increase, which is undesirable from the IPP's As seen from (13) and (14), we have relations for probability
point of view. This aspect of impelfect competition in the of acceptance with respect to the bid price and the expected
market and competitor behavior in response to their expectation market price. However, these functions are very general and
of the IPPs likely bids, should be considered by the IPP while certain practical restrictions need to be imposed so that they
formulating its bids. One way to respond to imperfect truly represent the probability trends expected of them. First,
competition is by re-adjusting c~1,to higher or lower values, as the range PAis restricted between 0 and 1. Also note that the
per requirements. This is modeled here by introducing an bidding range should be such that the probability should vary
adjustment factor to b,described by (12). from a maxima (at a certain bid, say zero bid for selling) to a
minima (say, infinite bid for selling) and vice versa.
With that concept, the bidding range is determined
as follows. The first derivative of the above functions gives
the point of inflexion (i.e. the minima or maxima). It is seen
that the point of inflexion, for both functions occur at 0 . 7 5 ~

110
and 1 . 2 5 ~This
. means that the probability of acceptance is
Revenue Earnings, M$ 1.367
maximum at a sell bid price of 0 . 7 5 ~and minimum at 1 . 2 5 ~ . Operation Cost, M$ 0.653
Similarly, a buy bid price of 1 . 2 5 ~will have maximum Profit, M$ 0.714
probability and 0 . 7 5 ~will have minimum probability of Generation, MWh . 74,220
acceptance. This can be explained by that the IPP does not Power Buy,MWh 2432
Power Sale, MWh 4686
risk bidding too high or too low to avoid rejection of the bid Bilateral Contracted Demand. MWh 65,423
or lowering of market prices. Using the relations in (9-14), a Reserve Energy, MWh 6542
simple search over a wide range of BPs would yield the bids,
which maximize P G ~ . 4.1 Strategic Bidding
As discussed in Section-3, the optimum bid is one with the
4. EXAMPLE highest PGA~,.PA, P+R and PGm are evaluated over the range,
The IEEE Reliability Test System [9] with 32 generators and [BPMIN, BPMAX] which is [0.75p, 1.25~1and that BP at
an installed capacity of 3405 MW, is used to represent an which P G A is ~ maximum, is identified. The bidding strategies
example IPP in the present studies. Details of the system are are developed for two cases, one for perfect competition in
given in the Appendix. The UC solution is obtained using the the market, i.e. assuming that the IPP is a price-taker and its
XAlO.0 MIP solver [lo] in GAMS (Generalized Algebraic bids have no influence on market price. The other case is
Modeling System) platform, a high-level programming imperfect competition in the market when the IPP’s bids may
language. The solver efficiently handles the present example affect the market price if it strategically bids too high or too
system UC problem with 32 generating units and 24 time- low. This is modeled in Equn.( 12) and discussed earlier.
intervals.
Table-3 shows the hourly operating and trading 4.1.1 Buy Bids
decisions of the IPP obtained from the solution of the price- Tables-5 and 6 shows the optimal bid prices for both, pelfect
based UC. The UC decisions for next 24 hours reveals that competition and an impelfect competition for each hour in
the IPP chooses to commit all its thermal units, even the which the utility has to BUY from the market. Based on the
costlier oil fired steam units. The trading decisions show that findings from Tables-5 and 6, the bidding strategies are
the IPP requires to buy power from the market during peak summarized for the buy bids in Table-7.
hours (9-14 and 17-20) in order to meet its bilateral contracts
while it sells power during the off-peak hours (2-8 and 21-
24). Table-4 shows a summary of the day’s operation.
--
E -
Ta
MC A
a
BP -
*
El
-
PA -P+R,:
1 43.5 22.7 28.13 1.o 1.o
able-3: IP 9 43.5 42.28 p < M C < pto 52.39 1.o 0.593 0.593
10 43.5 49.3 BPMlN < M C c p a 61.09 1.o 0.148
11 43.5 42.71 p < M C c pto 52.92 1.o 0.56
- 12 43.5 51.79 BPMlN c M C <p-o 64.17 1.o 0.054
1 2074 2074 0 383.53 2234.12 223.41 13 43.5 38.36 pto c M C c BPMAX 47.53 1a 0.891 0.891
2 3105 2531.88 221.07 0 2100.74 210.07 14 43.5 39.46 pto c M C e BPMAX 48.89 1.o 0.813 0.813
3 3105 2669.74 268.97 0 2000.70 200.07 17 43.5 41.26 p < MC c p+ cs 51.12 1.o 0.673 0.673
4 3105 2651.28 387.19 0 1967.36 196.74 18 43.5 40.68 p < M C e p+ ci 50.41 1.o 0.718 0.718
5 3105 2830.92 566.83 0 1967.36 196.74 19 43.5 53.40 p-ccMCcp 57.99 1.o 0.278
6 3105 3073.44 672.67 0 2000.70 200.07
7 3405 3405 690.72 0 2467.53 246.75 - 43.5 -
20 - 59.90 BPMIN > MC -
NIL - -
8 3405 3405 250.56 0 2867.67 286.77
9 3405 3389 0 95.55 3167.78 3 16.78
10 3405 3405 0 116.23 3201.12 320.1 1
11 3405 3389 0 132.23 3201.12 320.11
12 3405 3405 0 79.55 3167.78 316.78
13 3105 3089 0 395.55 3167.78 316.78
14 3105 3089 0 28.76 2834.33 283.43
15 3405 3405 360.6C 0 2767.64 276.76
16 3405 3389 307.92 0 2800.98 280.10
17 3405 3389 0 242.27 3301.16 330.12
18 3105 3089 0 578.95 3334.50 333.45
19 3405 3405 0 262.95 3334.50 333.45
20 3405 3405 0 116.23 3201.12 320.11
21 3405 3389 51.17 0.0 3034.40 303.44
22 3105 3068 23.60 0.0 2767.64 276.76 4.1.I . 1 Comments on the Bidding Strategiesfor BUY Bids
23 3105 2142.5 64.9 0.0 2434.19 243.42
Table-7 summarizes the bidding strategy for the buy bids in
-
24 3105 2530.93 -
220.1: 0.0 2100.74 210.07
six cases, depending on the MC and how it compares w.r.t p.
As seen from Table-7 and corresponding results listed in
Tables-5 and 6, PGm is very low for cases D, E and F. This
is because MC< p which means the IF’P’s own generation is
cheaper than the market price. Therefore, for any bid to incur

111
+R, the IPP has to bid very low (lower than the market price) the most favourable since the MC is very low and thus assures
but that will have very low probability of acceptance. The P + R = ~for
, any bid price. In this case therefore, the IPP must
other option is to bid high (BPMAX) for a high PA, but P+R bid at BPMIN to ensure PA=l and qualify for trading. This
will be very low. A bid just below MC is a good compromise kind of a case is usual for plants, which cannot be shut down
between the two because it would have a P + R =and ~ PA will (nuclear, for example) and hence the IPP bids very low into
be somewhat higher. If the IPP expects some market the market. In cases D and E, though P+R=l is not guaranteed,
imperfections, it should bid somewhat conservatively when it they have a high probability since p is very likely to be more
has to bid high. than MC. Therefore, the IPP should bid for high PA, too.
But overall, P G A ~is never higher than 0.2 or so, Therefore, the bids should be low. In case D, when it is
which means that there is only a 20% probability of the bid possible that the MC is quite close to p, the IPP can also
making any gains. Therefore the IPP may choose not to bid at choose to bid just over MC to ensure a P+R=l while seeing
all. This is obvious because when IPP's own marginal cost is that PAis within an acceptable range.
lower than the market price, it makes little sense to buy from Table
the market. However in imperfect market, PG" is somewhat
higher (but still considerably less) due to the consideration of
adjustment of p. 27.24 BPMIN < MC < p -a 20.43 1.O 0.993 0.993
17.20 BPMAX<MC NIL -
Cases-A, B and C are the most favourable cases for 22.76 p < MC < p + a 23.02 0.462 1.0 0.462
buyers since their MC is very high in these and their costly 32.06 BPMIN < MC < p -0 24.04 1.0 0.903 0.903
generation can be substituted by cheaper power purchased 44.84 pla<MC < p 43.52 0.595 1.0 0.595
from the market. Case A is the most favourable since it has 59.3 BPMIN>MC 44.47 1.0 1.0 1.0
the highest MC and will gain the maximum from purchase. 52.10 BPMIN < M C < p - o 39.08 1.0 0.955 0.955
42.55 p < M C < p + a 31.92 1.0 0.428 0.428
The IPP should bid high (BPMAX) in Case-A so as to 42.35 p < M C < p + a 31.76 1.0 0.412 0.412
increase its PA, while P+R is always =l. In cases B and C, the 30.04 p-o<MC< p 22.53 1.0 0.754 0.754
MC is still higher than expected market price, but very high 19.6 pto<MC<BPMAX 14.7 1.0 0.126 0.126
bids may lead to lower P+R. Hence, the IPP also has a choice -19.5 p+o<MC<BPMAX 14.63 1.0 0.112 0.112
to bid just below MC, which ensures P+R=l and still have a

T
Table
very high PA. sour MC P
If the IPP perceives that high bids (like BPMAX in 2 22.08 26.73
Case-A) could influence p by way of increasing it, which is 3 22.08 27.24 BPMIN < MC < p-la 22.09 0.832 1.0 0.832

,
against its interest in these cases as buyer, it should choose to 4 22.08 17.20 BPMAX <MC 12.9 0.852 0.237 0.202
5 23.0 22.76 p < MC < p + a 23.02 0.462 1.0 0.462
bid somewhat conservatively. Cases-B, C, 'being also 6 27.6 32.06 BPMIN<MC<pq 27.62 0.817 1.0 0.817
favourable for buyer, it can bid within the range which 7 43.5 44.84 P-CT< MC < p 43.52 0.595 I .O 0.595
ensures P+R=l while keeping PAwithin acceptable range. 8 43.5 59.3 BPMIN >MC 44.47 0.852 1.0 0.852
ategyfo;PowerBiJY pe 15 43.5 52.10 BPMIN < MC < p a 43.52 0.825 1.0 0.825
Bid in Perfect Bid in Imperfect 16 43.5 42.55 p < M C < p + a 43.51 0.427 1.0 0.427
Com tition Com tition 21 43.5 42.35 p < M C < p + a 43.51 0.412 1.0 0.412
BPMAX P+G
22 27.6 p -0< MC < p 27.62 0.752 1.0 0.752:
Just c MCIBPMAX (p+a)5 BP < MC 23 22.01 pw<MC<BPMAX 24.28 0.227 1.0 0.227
-
Just < MC I BPMAX (p 0)to MC 01 In
24 22.01 p+o<MC<BPMAX 24.16 0.227 1.0 0.227
(p+ a)to BPMAX impel Fect markets, the IPP should raise its b js
-
Just < MC I BPMAX (p a)to MC I
somewhat (overbid), when its MC is less than the expected
I NO BID (p+ 0)to BPMAX
I NO BID market price so as to prevent the system price from falling,
E 10,12 BPMIN < MC < p -0 BPMAX I NO BID Just < MCI NO BID which is against its interest as a seller.
F 20 BPMIN>MC NO BID BPMAX I NO BID For the cases A, B and C,the MC is higher than the
expected spot market price and the IPP should bid
4.1.2 Sell Bids strategically to maximize its gains, from these unfavourable
Tables-8 and 9 shows the optimal bid prices for both, perfect situations. Case A is the most unfavourable with a very high
and impelfect competition for those hours when the IPP has MC and P+Ris always very low. The IPP may choose not to
scheduled to SELL to the market. As earlier, the bidding bid at all. But if the market is imperfect and it can expect to
strategies are summarized for the sell bids as per the six cases influence the price to some extent, it should bid very low so
discussed in Section-4.1.l(Table-10). as to have a high P A .
Cases €3 and C are also unfavourable to sellers
4.1.2.1 Comments on the Bidding Strategies for SELL Bids though case-C is quite a close case, MC can be close to p, and
The bidding strategy for a seller is almost opposite to that of the bidder should be particularly cautious. Bids just over MC
the buyer. For the seller, the most favourable cases are D, E are the best options because they ensure P + p l while
and F when its marginal cost is less than the expected market maintaining PA within some range of acceptability.
price. In such cases it can sell its cheap power at a higher
market price and reap profits. Among D, E and F, case-F is

112
 CPk Bilateral contracted power selling price, $/MWh
GCsti Generating cost over and above minimum, $/MWh
EsVk Reserve generation available, MWh
(or NO BID) SDCi Shut-down cost of a generating unit, $
B 23,24 p +CT < MC < BPMAX Just > MC or BPMIN BPMAX STCi Start-up cost of a generating unit, $
(or NO BID) (or NO BID) pk Estimated spot-market price, $/MWh
C 5,16,21 p < M C < p + a Just>MCorBPMlN Just>MC Pk Actual spot-market price, $/MWh
(NO BID?) (NO BID?)
D 7,22, p -o< MC p Just > MC or BPMlN Just > MC
0 Standard deviation from p, $/MWh (assumed 0 . 1 ~ )
E 2,3,5, BPMIN<MC<pa BPMIN MC < BP< p-a Variables
15 BPk Bid price, $MWh
F 8 BPMIN>MC BPMIN BPMIN
~
PBk Power purchase by IPP from spot market, MWh
PGi,k Generation over and above minimum, MWh
Profit Profit of the IPP, $
MC IPP’s marginal cost of meeting its bilateral contracts
and scheduled trades, $MWh
PSI, Power sale by IPP to spot market, MWh
Ui.k Unit start-up decision (0: NO; 1: YES)
vi) Unit shut-down decision (0:NO; 1: YES)
Wi,k Unit ON/OJT status (0: O W , 1: ON)

7.2 System Description

Table-A.l: Unit Spechc Data of the IPP (Adopted from IEEE RTS [lo])
Unit Pmax, Pmin, GCst, Cmin STC SDC MUT,MDT,
MW MW WWh $ $ $ hrs hrs
1 (OFST) 12 2.40 27.60 15.88 5 1 1 1
2 (OFSTj 12 2.40 27.60 15.88 5 1 1 1
3 (OFST) 12 2.40 27.60 15.88 5 1 1 1
4 (OFST) 12 2.40 27.60 15.88 5 1 1 1
5 (OFST) 12 2.40 27.60 15.88 5 1 1 1
6 (CT) 20 16.0 43.50 15.00 6 2 1 1
7 (CT) 20 16.0 43.50 15.00 6 2 1 1
8 (CT) 20 16.0 43.50 15.00 6 2 1 1
9 (CT) 20 16.0 43.50 15.00 6 2 1 1
10 (CFST) 76 15.2 14.40 18.72 15 8 4 5
11 (CFST) 76 152 14.40 18.72 15 8 4 5
12 (CFST) 76 15.2 14.40 18.72 15 8 4 5
13 (CFST) 76 15.2 14.40 18.72 15 8 4 5
14 (OFST) 100 25.0 23.00 19.90 7 5 1 1
15 (OFST) 100 25.0 23.00 19.90 7 5 1 1
16 (OFST) 100 25.0 23.00 19.90 7 5 1 1
17 (CFST) 155 5425 11.64 13.44 35 10 4 5
18 (CFST) 155 54.25 11.64 13.44 35 10 4 5
19 (CFST) 155 54.25 11.64 13.44 35 10 4 5
20 (CFST) 155 5425 11.64 13.44 35 10 4 5
21 (OFST) 197 68.95 22.08 20.73 10 7 1 1
22 (OFST) 197 68.95 22.08 20.73 10 7 1 1
23 (OFST) 197 68.95 22.08 20.73 10 7 1 1
24 (CFST) 350 140.0 11.40 12.24 40 25 MR MR
25 (NUC) 400 100.0 6.00 7.53 45 25 MR MR
26 (NUC) 400 100.0 6.00 7.53 45 25 MR MR
27 (HYD) 50 0.0 2.0 0 0 0 1 1
28 (HYD) 50 0.0 2.0 0 0 0 1 1
29 (HYD) 50 0.0 2.0 0 0 0 1 1
30 (HYD) 50 0.0 2.0 0 0 0 1 1
31 (HYD) 50 0.0 2.0 0 0 0 1 1
32 (HY D) 50 0.0 2.0 0 0 0 1 1

Kankar Bhattacharya obtained his Ph.D. from Indian Institute of

Technology, New Delhi, in 1993. He was at Indira Gandhi Institute of
Development Research, Bombay, from 1993-98 and for short periods at
Intemational Institute for Applied Systems Analysis (USA). Austria and
Tampere University of Technology, Enland. Currently, he is with Chalmers
University of Technology, as assistant professor. His interests are in power
system dynamics, economic operations planning and power sector
deregulation.

113