SECURITY CONSTRAINED

ECONOMIC DESPATCH
of
Inter-state Generating Stations
pan-India

Detailed Feedback Report on

Expanded Pilot

March 2021

Power System Operation Corporation Limited

 TABLE OF CONTENTS

List of Figures ................................................................................................................ 4

List of Tables ................................................................................................................. 6
Executive Summary ...................................................................................................... 7
Disclaimer ....................................................................................................................10
Use of Information from this Report .......................................................................10
1 Introduction .............................................................................................................11
1.1 Background.............................................................................................................................................. 11
1.2 Regulatory Framework.......................................................................................................................... 11
1.3 Provisions in Draft IEGC 2020............................................................................................................. 12
1.4 Regulatory Framework for Real-Time Market ................................................................................. 13
2 Implementation of Expanded SCED Pilot ............................................................. 14
2.1 Regulatory Provisions ............................................................................................................................ 14
2.1.1 Extending the period of SCED pilot .......................................................................................... 14
2.1.2 Expanding the scope of SCED pilot........................................................................................... 14
2.1.3 Participation in the pilot............................................................................................................... 14
2.1.4 Declaration of Variable Charges .............................................................................................. 15
2.1.5 Declaration of Technical Parameters ........................................................................................ 15
2.1.6 Sharing of System Savings .......................................................................................................... 16
2.1.7 Scheduling Process under RTM and SCED ................................................................................ 18
2.2 Expanding Ambit of SCED participating generators ...................................................................... 18
2.3 Challenges faced in implementation of expanded SCED pilot .................................................... 19
2.3.1 Declaration of Variable Charges by Generating Stations having part-tied capacity .. 20
2.3.2 Treatment of capacity without having any Access in transmission system ......................... 21
2.3.3 Heat Rate Compensation ............................................................................................................. 22
2.3.4 Coal Linkage .................................................................................................................................. 22
3 SCED Development & Operation in 2020 ............................................................. 23
3.1 Implementation of RTM ......................................................................................................................... 23
3.2 New Scheduling Timelines..................................................................................................................... 23
3.3 Higher ramp rates in thermal ISGS .................................................................................................... 25
3.4 Impact of COVID-19 pandemic .......................................................................................................... 26
3.5 New features in SCED ........................................................................................................................... 28
3.5.1 Modular Integration with RRAS .................................................................................................. 28
3.5.2 Incorporation of Export ATC ....................................................................................................... 30
3.5.3 Soft landing on exclusion ............................................................................................................. 30
3.5.4 Scheduled and Emergency Stop ................................................................................................ 31
3.5.5 Evacuation Limit ............................................................................................................................. 33
3.6 Reduction in Costs ................................................................................................................................... 33

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 1

 3.7 Ease of generator operations .............................................................................................................. 40
3.8 Variations in Variable Cost .................................................................................................................. 42
4 Look Ahead SCED ....................................................................................................47
4.1 Infeasibilities............................................................................................................................................ 48
4.2 Cost Reduction......................................................................................................................................... 49
4.3 RRAS Dispatch ......................................................................................................................................... 50
4.4 Availability of Ramp-constrained reserves ....................................................................................... 51
4.5 Ease of generator operations .............................................................................................................. 52
4.6 Schedule variability within look-ahead window .............................................................................. 53
5 Security Constrained Unit Commitment ............................................................... 57
5.1 Motivation ................................................................................................................................................ 57
5.2 SCUC Formulation .................................................................................................................................. 59
5.2.1 Objective function ......................................................................................................................... 59
5.2.2 Constraints ...................................................................................................................................... 59
5.3 Dovetailing SCUC to the existing framework ................................................................................... 61
5.4 Changes to Web Based Scheduling Software (WBES) .................................................................. 62
5.4.1 Unit DC ............................................................................................................................................ 62
5.4.2 Pmax in WBES and SCED ............................................................................................................ 63
5.4.3 Pmin in WBES and SCED.............................................................................................................. 63
5.4.4 Requisitions by Constituents ......................................................................................................... 63
5.4.5 Unit De-Commitment through SCUC .......................................................................................... 63
5.4.6 Unit Commitment through SCUC ................................................................................................. 64
5.5 SCUC Case Study................................................................................................................................... 64
5.5.1 Inputs and Assumptions................................................................................................................. 64
5.5.2 Classification of Units in SCUC ................................................................................................... 65
5.5.3 Impact on Reserves ....................................................................................................................... 68
5.5.4 Change in schedule due to SCUC+SCED ................................................................................. 72
5.5.5 Unit Status ....................................................................................................................................... 76
5.5.6 Impact of SCUC on startups ........................................................................................................ 78
5.6 Optimization Formulation of SCUC..................................................................................................... 79
5.6.1 Notation .......................................................................................................................................... 80
5.6.2 Objective function ......................................................................................................................... 81
5.6.3 Startup and Shutdown Variables............................................................................................... 82
5.6.4 Constraints ...................................................................................................................................... 83
5.7 Compensation for heat rate degradation......................................................................................... 88
5.8 Startup Costs due to SCUC .................................................................................................................. 89
5.9 Settlement Related Issues under SCUC .............................................................................................. 89
6 Way Forward ............................................................................................................91
6.1 Extension of the SCED Pilot Project ..................................................................................................... 91
6.2 Need for SCUC pilot project ............................................................................................................... 91

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 2

 6.3 Increase Frequency of Declaration of Variable Charges .............................................................. 92
6.4 Declaration of Incremental Heat Rate Curves .................................................................................. 93
6.5 Need to bring down Minimum Turn-Down Level .............................................................................. 93
6.6 Spinning (Hot) and Non-Spinning (Cold) Reserves........................................................................... 94
6.7 National Pool Account ........................................................................................................................... 94
6.8 Statement of Station Heat Rate Degradation Compensation ....................................................... 94
6.9 Replication of SCED at Intra State level through SLDCs ................................................................ 95
6.10 Optimization considering full transmission network ......................................................................... 95
Annexures ....................................................................................................................96

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 3

 LIST OF FIGURES
Figure 1 Composite Timeline for dispatch of SCED and RRAS (w.e.f. 19th September 2020) ................ 24
Figure 2 Block-wise ramp in post-SCED schedule of Talcher-II (Jan-Dec 2020) ......................................... 25
Figure 3 Average Net Schedule of SCED generators & All India Energy met ............................................ 27
Figure 4 Distribution of SCED SMP ....................................................................................................................... 27
Figure 5 Architectural overview of RRAS implementation ................................................................................ 29
Figure 6 Soft landing on exclusion ........................................................................................................................ 31
Figure 7 Scheduled and Emergency Stop in SCED ............................................................................................ 32
Figure 8 Daily cost reduction due to SCED (Jan 2020-Jan 2021) ................................................................. 34
Figure 9 Daily Average Perturbation in SCED ................................................................................................... 35
Figure 10 Average SCED Perturbation vs Daily Cost Reduction .................................................................... 35
Figure 11 Daily Sale Volume in RTM by SCED generators ............................................................................. 36
Figure 12 Impact of RTM Trades on SCED Cost Reduction – study results ................................................... 37
Figure 13 Perturbation by SCED: Dec 2019 vs Dec 2020 .............................................................................. 37
Figure 14 Month-wise box plot of Variable Charges of SCED generators ................................................. 46
Figure 15 Trend of daily average cost of generation under SCED............................................................... 46
Figure 16 Rolling look-ahead optimization window ......................................................................................... 48
Figure 17 Net Infeasibilities in SCED – Single period vs. Look-ahead.......................................................... 49
Figure 18 Cost reduction on sample days – Single period vs Look-ahead.................................................. 50
Figure 19 RRAS dispatch on sample days – Single period vs Look-ahead ................................................. 50
Figure 20 Ramp constrained reserves on sample day – Single period vs Look-ahead ............................ 51
Figure 21 Duration curve of ramp constrained reserves over study period ................................................ 51
Figure 22 Ease of generator operations with SCED – Single period vs Look-ahead ................................ 52
Figure 23 Change in All India ISGS schedule over day ................................................................................... 53
Figure 24 Net ISGS schedule within look-ahead run window ......................................................................... 54
Figure 25 Variability in block total schedule visibility over time.................................................................... 55
Figure 26 Variation of block total ISGS schedule across runs ........................................................................ 56
Figure 27 Regulation quantum duration available during 2019-20 (MW) ................................................. 58
Figure 28 Surplus of sell bids over buy bids in Day-Ahead Market in Power Exchange ......................... 59
Figure 29 Classification of units post SCUC ........................................................................................................ 66
Figure 30 Spinning Reserves comparison from 01 Jun to 03 Jun ................................................................... 67
Figure 31 Seasonal Transition of Reserves.......................................................................................................... 68
Figure 32 Ramp-constrained Reserves comparison ........................................................................................... 69
Figure 33 Spinning reserves comparison ............................................................................................................. 70
Figure 34 SMP Heat Map ...................................................................................................................................... 71

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 4

Figure 35 SMP vs Cheapest Cold Reserve .......................................................................................................... 71
Figure 36 Daily additional savings due to SCUC and startup costs .............................................................. 72
Figure 37 All India SCUC Schedule vs Actual Schedule ................................................................................... 73
Figure 38 ER SCUC Schedule vs Actual Schedule .............................................................................................. 73
Figure 39 NER SCUC Schedule vs Actual Schedule ........................................................................................... 73
Figure 40 NR SCUC Schedule vs Actual Schedule ............................................................................................. 74
Figure 41 SR SCUC Schedule vs Actual Schedule .............................................................................................. 74
Figure 42 WR SCUC Schedule vs Actual Schedule............................................................................................ 74
Figure 43 IGSTPS Jhajjar SCUC Schedule vs Actual Schedule ....................................................................... 75
Figure 44 Dadri-II SCUC Schedule vs Actual Schedule..................................................................................... 75
Figure 45 Solapur SCUC Schedule vs Actual Schedule .................................................................................... 75
Figure 46 Mauda-II SCUC Schedule vs Actual Schedule .................................................................................. 76
Figure 47 Unit status comparison for June 2020 to August 2020.................................................................. 77
Figure 48 Unit status comparison for December 2020 to January 2021..................................................... 78
Figure 49 Startups Scatter plot ............................................................................................................................. 79
Figure 50 Diurnal startup pattern ......................................................................................................................... 79
Figure 51 Startup and Shutdown variables ........................................................................................................ 83
Figure 52 Illustration of Minimum up time (a)Valid (b)Invalid ......................................................................... 85
Figure 53 Illustration of Minimum up time (a)Valid (b)Invalid ......................................................................... 86
Figure 54 Illustration of Minimum down time (a)Valid (b)Invalid.................................................................... 87
Figure 55 Illustration of Minimum down time (a)Valid (b)Invalid .................................................................... 87
Figure 56 Illustration of crew constraint ............................................................................................................... 88

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 5

 LIST OF TABLES
Table 1 Generator wise SCED statement (Apr 2019 – Jan 2021) ............................................................... 38
Table 2 Generator-wise change in number of instructions and Delta MW .................................................. 40
Table 3 Distribution of Variable Charges of SCED Generators (Apr-19 onwards) ................................... 43
Table 4 Month-on-month change in Variable Charges (Apr-2019 onwards).............................................. 44
Table 5 Difference between VC of successive generators in SCED merit stack (Apr-19 onwards) ........ 45
Table 6 Inputs required for SCUC ........................................................................................................................ 61
Table 7 Proposed scheduling activities and timelines for SCUC..................................................................... 61
Table 8 Cheapest and Costliest unit within each group ................................................................................... 66
Table 9 Units impacted by SCUC (Quadrants 2 and 4) .................................................................................. 66
Table 10 SCUC Cost Summary (Jun 2020 – Jan 2021) .................................................................................. 71

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 6

 EXECUTIVE SUMMARY
The pilot on SCED in ISGS pan-India has been implemented since 1st April, 2019. The pilot
helped to explore the scope of optimization on pan-India basis. The total production costs in
ISGS are minimized without major structural changes in the prevailing scheduling and
settlement system. Central Commission, on 18th April, 2020, directed the extension and
expansion of the pilot on SCED till 31st March, 2021. A detailed feedback report on the pilot
was submitted to the Commission in January 2020. The present report would cover the
experience of the pilot and feedback of activities during the extended & expanded pilot.

During the period of the expanded pilot, the SCED application was further developed. The
generating stations with part tied-up capacity and merchant generators were also included.
RTM was operationalized entailing changes in scheduling timelines. Revamped optimization-
centric RRAS software was implemented. The formulations of Look Ahead SCED and Security
Constrained Unit Commitment were also tested as part of pilot project.

SCED application has been continuously improved over time with addition of new features.
With growth in renewables, the inter-regional flow patterns are undergoing a seasonal and
diurnal shift. In view of this shift, the SCED software has been modified to incorporate import
as well as export ATC constraints on inter-regional schedules for all regions. A soft-landing
feature has been added to SCED software, which ensures that generators return to their pre-
SCED schedules following ramp rates on exclusion. The SCED stop functionality was split into
two types depending on the nature of event requiring SCED to be stopped – Scheduled Stop
and Emergency Stop. During contingencies in the grid operations, in order to facilitate quicker
and more effective control of generation, the evacuation limit feature has also been added to
SCED.

1600 MW of additional generation capacity was added under the expanded pilot viz. Farakka-
III (500 MW), Korba-III (500 MW) and Jhabua (600 MW). In order to enable participation of
merchant power plants, the sharing of required information by generators needed for SCED in
scheduling interface (WBES) of RLDC was enabled. The required changes/modification in SCED
software to accommodate the merchant power plant were also carried out.

Most of the partly-tied generators have signed PPAs with different basis of variable charges.
The challenge of declaration of variable charges by generating stations having part-tied

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 7

capacity was faced. The weighted average PPA rate of the generator was considered for the
pilot which was also be applicable for the un-tied/merchant quantum.

The treatment of generation without having any access in ISTS was also a challenge. In order
to inject power in ISTS, it is mandatory to obtain some form of access i.e., long-term, medium-
term or short-term. SCED schedules can, thus, be given for a maximum quantum for which
long-term, medium-term or short-term access has been obtained by the concerned
generating station.

The operation below the normative PLF (85%) would entail heat rate degradation
compensation in accordance with PPA provisions and IEGC. Partly-tied and merchant
generators with verified PPAs having provisions of heat rate degradation compensation would
be eligible for compensation.

With RTM in operation, RRAS and SCED dispatch is taking place after RTM and the URS
available post-RTM and RRAS is used for SCED optimization. In order to facilitate quick and
automatic incorporation of RTM transactions in schedules, an in-house software application
was developed to provide an interface to power exchanges and WBES along with verification
of margins and clearing of transactions. The faster dissemination of schedule data for bidding
decision in RTM has been facilitated through a secure API shared with generators and
constituents. With the implementation of RTM, new scheduling timelines came into vogue with
the gate closure for DC and requisition revisions getting shifted from 4th time block to 7th/8th
time block before dispatch block.

The pan-India lockdown for containment of COVID-19 pandemic from 25th March 2020
onwards had an impact with demand reduction up to 25-30 %. A significant portion of
generation capacity went under reserve shutdown. There was corresponding reduction in DC
on-bar and net schedule of SCED participating generators. There was also an impact in SCED
System Marginal Price due to demand reduction.

During 2020 (Jan-Dec), the total reduction in variable cost of generation due to SCED was ₹
752 Crores. From the start of the SCED pilot in April 2019 up to January 2021, the cumulative
savings in generation cost on All-India basis is ₹ 1624 crores.

With the optimal fuel supply arrangements and rationalization of coal stock, the spread of
variable charges is reducing pan-India. The reduction of spread of variable charges appears to

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 8

be one of the causes for lowering of SCED savings in 2020 vis-a-vis 2019.

A look-ahead SCED algorithm has been developed and its operation and results have been
simulated for a period of one week. The key driver for a look-ahead SCED implementation is
reduction of potential infeasibilities as the model has a longer term visibility. It was observed
that for around 10% of time during the simulated week, ramping up reserves are zero even
with the look-ahead model. This is owing to exhausted spinning reserves in ISGS. Introduction
of unit commitment with SCUC would help in maintaining sufficient spinning reserves.

An in-house Security Constrained Unit Commitment (SCUC) software has been built with the
objective to minimize the operation cost while also maintaining requisite reserves for a rolling
3 day window, in-line with draft IEGC 2020. The operation costs comprise of energy costs and
startup costs.

The spinning reserve would be required to be maintained to take care of the real time
uncertainties like unit tripping, unit derating, errors in demand and RE forecast, and weather
events.

ISGS SCUC case study was run for an eight-month period from June 1, 2020 to Jan 31, 2021
on a 3-day rolling window basis. The different aspects of reserves, unit status and changes in
schedule during the case study period have been detailed in the report.

The changes to the existing scheduling activities and timelines for SCUC have also been
detailed in the report. Continuation of SCED and operationalization of SCUC with reserves
appear to be the next logical interventions that can be taken up for ensuring security and
economy in the Indian Power System.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 9

 DISCLAIMER

Precautions have been taken by Power System Operation Corporation Ltd.

(POSOCO) to ensure the accuracy of data / information and the data / information in

this report is believed to be accurate, reliable and complete. However, before relying
on the information material from this report, users are advised to independently

ensure its accuracy, currency, completeness and relevance for their purposes, and, in
this respect, POSOCO shall not be responsible for any errors or omissions. All

information is provided without warranty of any kind.

POSOCO disclaims all express, implied, and statutory warranties of any kind to

user and/or any third party, including warranties as to accuracy, timeliness,

completeness, merchantability, or fitness for any particular purpose. POSOCO have no

liability in tort, contract, or otherwise to user and/or third party. Further, POSOCO shall,

under no circumstances, be liable to user, and/or any third party, for any lost profits or
lost opportunity, indirect, special, consequential, incidental, or punitive damages

whatsoever, even if POSOCO has been advised of the possibility of such damages.

By reading this report, the user/reader confirms about their awareness and agreement

to this disclaimer and associated terms referred elsewhere.

USE OF INFORMATION FROM THIS REPORT

This report has been compiled by Power System Operation Corporation Ltd.

(POSOCO) and is a result of contribution in some or other form of many experts,

engineers, power system professionals, data from the reports published by statutory

authorities, data from the web-sites and internal software applications. This report is
being published for the regulatory compliance. Information in this report can be used

without obtaining the permission of POSOCO. However, the material used from this
report should duly acknowledge by way of citing the name of report and source as
POSOCO.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 10

 1 INTRODUCTION
1.1 Background
The implementation of a Pilot on SCED in ISGS Pan India was directed by Central

Electricity Regulatory Commission i.e. Central Commission w.e.f 1st April, 20191 for six
months which was extended upto 31st March, 20202 and further upto 31st May, 20203.

The prime driver behind the pilot on SCED was to explore the scope of optimization
and, therefore, the possibility of minimising the total production costs without major

structural changes in the prevalent system. Central Commission, on 18 th April, 20204,

directed the extension and expansion of the pilot on SCED till 31st March, 2021.

During the period of the expanded pilot, new features in SCED application were
added. The generating stations with part tied-up capacity and merchant generators

were also included. Real Time Market (RTM)5 was operationalized entailing changes in
scheduling timelines. The formulations of Look Ahead SCED and Security Constrained

Unit Commitment were also tested as part of pilot project.

1.2 Regulatory Framework

As per the Central Commission direction, during the initial phase till May 2020, the
pilot on SCED was implemented for the limited set of the thermal ISGS that are regional
entities and whose tariff is determined or adopted by the Central Commission for their
full capacity honouring the existing scheduling practices prescribed in the Grid Code.

It was recognized that there would still be scope of optimisation through SCED after
the gate closure which has been implemented as enabler for RTM.

Since June, 2020, the participation in the expanded pilot has been opened to all
willing generators pan-India by including generating stations with part tied-up

capacity, merchant generators and intra-state generators in addition to already

1 http://cercind.gov.in/2019/orders/02-SM-2019.pdf
2 http://cercind.gov.in/2019/orders/08-SM-2019.pdf
3 http://cercind.gov.in/2020/orders/Suo-Motu_1-SM-2020.pdf
4 http://cercind.gov.in/2020/orders/8-SM-2020.pdf
5 http://cercind.gov.in/2019/regulation/1.%20Statement%20of%20Reasons_RTM_12_12_2019.pdf

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 11

participating generators. Upfront declaration of variable charges along with technical,
operational and other parameters to the system operator has been mandated by the

Central Commission. Further, new methodology for sharing of benefits amongst the
beneficiaries and the participating generators has been put in place.

1.3 Provisions in Draft IEGC 2020

The Central Commission constituted an Expert Group to review Indian Electricity

Grid Code and other related issues in May, 2019. The expert group submitted the
report in January, 20206. The draft revised IEGC proposed by the expert group has the

provisions to ensure and monitor availability of reserve capacity. The quantum of

reserve capacity required to be maintained for grid security is related to credible

contingency including net error in the forecasts of demand and renewable generation.

It is proposed that SLDC/RLDC may calculate the desired secondary reserve to

be kept in their control areas at the beginning of each financial year and submit to

NLDC. The quantum of reserves earmarked for secondary would be based on the data
of last year Area Control Error (ACE), will be taking care of exceptional high values due

to weather related phenomenon or any other exceptional circumstances. It is desirable

that reserves may be provided locally by the control area.

The responsibility to provide reserve response may be shared by all Control

Areas in a distributed manner in the interest of grid security and in a participative

manner so that the frequency is collectively controlled through ACE minimization.

Tertiary reserves may be arranged from the generating stations, Energy Storage

Systems (ESS) and/or through demand response. Tertiary reserve has to be greater or
equal to secondary reserves to take care of contingencies, and would be maintained

at both regional entity level as well as state control area. Tertiary reserves will act as
replenishment for secondary reserves as secondary reserves are to be restored back to

their original level for preparation towards next contingency.

6 http://cercind.gov.in/2020/reports/Final%20Report%20dated%2014.1.2020.pdf

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 12

 It has been proposed that the Security Constrained Unit Commitment (SCUC)
exercise may be carried out to facilitate reliability of supply to the regional

entities/beneficiaries taking into account optimal cost, adequate reserves, ramping

requirements factoring security constraints. In order to ensure availability of adequate

secondary and tertiary reserves with sufficient ramping capability, NLDC may identify
the generating unit for purpose of unit commitment at the national level for regional

entity generating stations on a 3-day rolling basis. NLDC, through RLDC may advise
the regional entity generators to commit or de-commit the unit.

1.4 Regulatory Framework for Real-Time Market

RTM has been implemented with effect from 1st June 2020. The Real Time

Market of electricity is an organized market platform which focuses on bringing in vital

flexibility to manage real time balance and at the same time it also facilitates optimal

utilization of the available surplus generation in the system. RTM for electricity has

been intended for thirty minutes delivery period in two consecutives fifteen-minute
time blocks at regular intervals during the day which implies RTM has 48 sessions in a

day starting from 00:00 hrs up to 24:00 hrs. The mechanism for price discovery has
been adopted as double-sided closed auction with uniform price discovery. Members

participating in the Real Time Market have the choice of setting buy and sell offers for
each fifteen-minute time block in the half hourly Real Time Market. To bring in firmness

in the schedule and transactions in real time, Gate Closure has been introduced.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 13

 2 IMPLEMENTATION OF EXPANDED SCED PILOT

2.1 Regulatory Provisions

2.1.1 Extending the period of SCED pilot

The Central Commission recognized that the scope of optimisation and, therefore,
the possibility of minimising system cost steers the need for appropriate regulatory

framework in the context. The optimization of the last-mile available resources through
a RTM market platform was also envisaged. The scope of optimisation through SCED

after the gate closure was considered. Accordingly, the Central Commission directed
to extend the SCED pilot beyond 1st June, 2020, for a further period up to 31st March,

2021.

2.1.2 Expanding the scope of SCED pilot

The Central Commission also decided to expand the ambit of SCED by including
the generators other than the regulated thermal ISGS. During the extended period, the

SCED pilot would be open to all generating stations that are willing to participate
during the extended period upto 31st March, 2021. These include generating stations

owned by the State entity and having capabilities to communicate with RLDCs/ NLDC;
generating stations whose scheduling is done by RLDCs; and State-embedded

generating stations whose scheduling, metering, accounting and settlement is in place

and whose scheduling related information exchange can be enabled through the SLDC

interfacing with the concerned RLDC/ NLDC. Scheduling for those generators that are
done through SLDC, shall continue to be done by the respective SLDC while increment/

decrement instructions under SCED shall be communicated from NLDC/ RLDCs to the
respective SLDC.

2.1.3 Participation in the pilot

Generators already participating in the SCED pilot would continue to participate in

the Pilot for the extended period. Other generators willing to participate in the
extended period are required to provide a one-time consent for participation in SCED

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 14

pilot. Once the consent is communicated to POSOCO, it is mandatory for the generator
to participate in the SCED pilot for a minimum period of one month. Such generators

would declare their variable charge upfront, similar to the existing SCED generators
participating in the pilot.

2.1.4 Declaration of Variable Charges

Variable charge of generators other than the ISGS whose tariff is determined by

the Commission is considered as under:

a) For generators having full capacity tied in multiple PPAs, variable charge shall be

considered as the weighted average of the variable charges, as determined or

adopted by the Appropriate Commission.

b) For generators having only part capacity tied in PPA(s), variable charge shall be

considered as the weighted average of the variable charges, as determined or

adopted by the Appropriate Commission for the contracted capacity shall be

considered for the untied portion of the capacity as well.

c) For generators with no tied capacity (merchant generator), variable charge shall

be considered as intimated by such generator at least one week in advance to

POSOCO. The generators shall be required to declare the variable charge on

monthly basis.

2.1.5 Declaration of Technical Parameters

Generators participating in the pilot shall also be required to declare on monthly

basis relevant technical parameters as required by NLDC/ RLDCs/RPCs, including but

not limited to installed capacity, declared capability (DC), Technical Minimum, Ramp
up/ Ramp down capability etc. Generators shall also communicate the details of

Discom/ constituent-wise share in generating station, requisition (day-ahead and last

revision) from the generating station to respective RLDC/ SLDC.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 15

 2.1.6 Sharing of System Savings

The Central Commission decided to bring parity for sharing the net savings as a

result of SCED during the extended period with the benefit sharing mechanism
(sharing of net gains) specified for Real Time Market (RTM) in respect of tied capacity

of generators. It is to address the possibility of any arbitrage between RTM and SCED
for the participating generators, as a generator, not getting cleared in RTM, has the

option of participating in SCED.

Accordingly, the Commission directed that the net savings as a result of SCED after

adjusting heat rate compensation for part load operation of the generators shall be

shared in the following manner:

a. As a first step, the share towards ‘untied capacity’ of merchant generators as

well as generators with part capacity tied would be segregated from the net
benefits, in the ratio of contribution of such generators to SCED, for every time

block.
b. The remaining benefits (after segregation as at step ‘a’ above) are then shared

in the ratio of 50:50 between the generators (with tied capacity, participating in
SCED) and the concerned beneficiaries/ Discoms, aggregated on a monthly

basis as per Regional Energy Account (REA) and weekly SCED accounts. The
benefits (50% of the remaining benefits after segregation as at step ‘a’ above)

corresponding to the beneficiaries/Discoms are shared in proportion to their

final schedule from the generating stations covered under SCED pilot.

c. The benefit of generators with tied capacity (i.e. 50% of the remaining benefits
after segregation as at step ‘a’ above) are shared between SCED Up and SCED

Down generators in the ratio of 60:40 for respective time block.

d. Based on the above, if a generator’s share exceeds 7 paise/kWh the same is

restricted to a ceiling of 7 paise/kWh and the gains over and above 7 paise/kWh
would be shared among Discoms in the same manner as at clause ‘b’ above.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 16

 e. Accordingly, the benefits for each time block are summed monthly for each
generator participating in the SCED pilot, as per the Regional Energy Account

(REA). This requires that all accounts including heat rate compensation for part
load operation are issued in a timely manner so as to facilitate smooth

disbursement and settlement.

f. The cap of 7 paise/kWh is, however, not be applicable in respect of ‘untied

capacity’ of merchant generators as well as generators with part capacity tied,

for its untied capacity. The share of savings from SCED for generators with part

capacity tied, is first settled in proportion to their contribution from tied capacity

and then from untied capacities. The share in savings as segregated at step ‘a’
above in respect of such generators is same for SCED up and SCED down i.e.

the ratio of 60:40 as in case of generators with tied up capacity will not be
applicable in this case.

Hence, the allocations for merchant generators as segregated in the step at ‘a’
above, would accrue to such generators as their final share. Similarly, for generators

with part capacity tied, the allocation, as segregated in the first step at (a) above,
proportionate to the untied capacity, would accrue to such generators as their final

share for the untied capacity.

The above sharing mechanism is subject to the following:

a. For Inter State Generating Stations (ISGS) that are regional entities and whose
tariff is determined or adopted by the Commission, the heat rate compensation

for part load operation shall be provided as per CERC (Indian Electricity Grid
Code) Regulations, 2010 as amended from time to time.

b. For generators with part capacity tied (whose PPAs have been approved by the
Appropriate Commission), the heat rate compensation for part load operation

shall be provided as per CERC (Indian Electricity Grid Code) Regulations, 2010
as amended from time to time. The Station Heat Rate (SHR) to be used as

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 17

 reference for this purpose shall be according to the provisions in the respective
PPA(s) read with the relevant regulations of the Appropriate Commission.

c. Generators with no tied capacities (merchant generators) shall not be eligible

for any heat rate compensation for part load operation.

2.1.7 Scheduling Process under RTM and SCED

There was need for further streamlining the scheduling process, especially with the

introduction of RTM. The half-hourly RTM brought the required flexibility in the system
and the introduction of gate closure provided more certainty in despatch, especially in

terms of reserve requirement. The stakeholders would use the RTM platform – the

buyers for meeting their real time energy needs and the sellers, including the
generators having un-requisitioned surplus (URS), for selling their schedulable surplus.

SCED is another window of optimisation for a particular time block after the RTM
window for that time block has closed. The Central Commission directed to run SCED

optimization after gate closure.

2.2 Expanding Ambit of SCED participating generators

The following types of generators have also been allowed to participate in the
expanded SCED pilot:

(a) Regional entity generator

a. Power station with part capacity tied (whose PPAs have been approved

by the Appropriate Commission).

b. Power station with no tied capacities (merchant generators)

(b) State generator

a. Owned by the State entity.

b. State-embedded generating stations.

In case of state owned/embedded generator whose scheduling, metering,

accounting and settlement is in place and whose scheduling related information

exchange can be enabled through the SLDC interfacing with the concerned RLDC/

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 18

NLDC are allowed to participate. Based on the guidelines, an analysis was carried out
to identify the prospective regional entity generators who can participle under SCED.

In order to give an overview of SCED mechanism to regional entity generators, two

e-meetings were conducted with stakeholders. One meeting was with Southern

Region on 6th July, 2020 and another meeting was with Western Region on 8 th July,
2020. The meeting was attended by the regional entity generators, RPCs, RLDCs and

NLDC.

Extensive deliberation/discussion were held to clarify the query of regional entity

generators and RPCs. The meetings were also helpful in understanding the various

concerns of the thermal power stations. A detailed report in this regard was submitted
by NLDC on 26th August, 2020 to the Central Commission (Annexure-1).

The capacity added under Expanded SCED pilot is as follows:

Date from (Inclusion Date to (Excluded

Station Name Company Region
in SCED) from SCED)
Farakka-III
NTPC ER 16th June 2020
(500 MW)
Korba-III
NTPC WR 16th June 2020
(500 MW)
Jhabua Avantha 17th August 2020 16th September 2020
WR
(600 MW) Power 16th November 2020 16th December 2020

2.3 Challenges faced in implementation of expanded SCED

pilot

In order to enable participation of merchant power plants such as Jhabua Power

Ltd., the required information needed for SCED in Web Based Energy Scheduling

software (WBES) of RLDC was shared by the generator. The necessary changes were
deployed in WBES including an application programming interface software for

interfacing with the SCED software at NLDC. In order to get acquainted with the
process of submitting the Pmax, Pmin, Ramp Up, Ramp Down details in WBES, mock

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 19

data submission was also carried out. The required changes/modification in SCED
software (In-house developed by POSOCO) to accommodate the merchant power

plant were to be carried out in an expeditious manner.

2.3.1 Declaration of Variable Charges by Generating Stations having part-

tied capacity

Most of the partly-tied generators have signed PPAs with different basis of variable

charges. Some even have a single part tariff PPAs and in some cases, water charges,
EDC charges etc. are also included in variable charges. The variable charge mentioned

in PPAs is not reflective of actual variable charge as they were signed with different

assumptions. In case of under-recovery of variable charges, fixed cost components was

proportionately increased as pass through in PPA.

In case of JPL, as per PPAs signed with Kerala and West Bengal under Section 63,
all the Auxiliary Power consumption (APC) and water charges are part of Fixed Cost.

However, in case of PPA signed with Madhya Pradesh under Section 62 in line with
CERC/MPSERC Regulations, both APC and water charges, which are on reimbursable

basis, are considered under VC. As per the CERC Terms and Conditions of Tariff
Regulations, water charges are part of O&M. For the purpose of SCED, it was clarified

to Jhabua, that the weighted average variable charges declared by the station should
be as per the billing for the previous month. There would be no segregation of variable

charges for a single plant. During the deliberations, Jhabua Power Ltd. agreed that it
would provide weighted average PPA rate for SCED and this would also be applicable

for the un-tied/merchant quantum.

It was highlighted that some generators declare VC at ex-bus and some generators

declare at beneficiary boundary. It was clarified that the variable charges would be
considered as per the CERC order. It was highlighted that consideration of violation

penalties and deterioration in heat rate, the real picture of VC would come. It was
clarified that to begin with, the simplified VC is being taken. In future, there would be
need for incremental heat rate curve for every generator.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 20

 2.3.2 Treatment of capacity without having any Access in transmission
system

As per the SCED extension order, power plant with part merchant or full merchant
capacity is allowed to participate under SCED. The power station participating under

SCED may be required to generate up to their maximum declared capacity (Pmax

which includes both tied & untied/merchant portion) depending on the system

marginal price vis-à-vis the variable charge declared by the concerned generating
station. Power station with part merchant or full merchant capacity may be required

to inject in grid including merchant quantum which was left after all opportunities for

sale of power in the market have been exhausted.

As per the Regulation 8.6 of the CERC Grant of Connectivity, Long-term Access and

Medium-term Open Access in inter-State Transmission and related matters

Regulations, in order to inject power in grid, it is mandatory to obtain some form of

access i.e., long-term, medium-term or short-term. The relevant extract is given below.

“The grant of connectivity shall not entitle an applicant to interchange any power

with the grid unless it obtains long-term access, medium-term open access or short
term open access”

Further, the Regional Transmission Account (RTA), prepared in accordance with the
extant POC Regulations, considers LTA+MTOA+Allocation (deemed LTA) quantum as

the approved injection limit.

Similar issues may arise in case of State Generator participating under SCED. They

may not have access to inject in ISTS network, but they may get injection schedule
under SCED to ISTS. In order to comply with the above mentioned regulatory

provisions, SCED schedules can thus be given for a maximum quantum for which long-
term, medium-term or short-term access has been obtained by the concerned

generating station. In this regard, schedules under SCED can also be given if the plant
has taken access to a target region without specifying the buyer. Hence, in the case of

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 21

Jhabua, to start with, Pmax for SCED has been considered as the Approved
LTA+MTOA+STOA+PX+RTM quantum only.

2.3.3 Heat Rate Compensation

The declaration of the variable charge is on the basis of normative PLF of 85%. The

operation below the normative PLF would entail heat rate degradation compensation
in accordance with PPA provisions and Grid Code. The partly-tied and merchant

generators may have to verify whether the respective PPAs have provisions of heat
rate degradation compensation. If there is no provision, no compensation would be

made.

2.3.4 Coal Linkage

In case of part tied-up generator, it has been clarified that linkage coal has to be

used for tied capacity and non-linkage coal for untied capacity. The generator has to
perform internal due diligence regarding viability. There is no one to one matching

done in the SCED optimization. The declaration of DC by station is based on fuel

availability and the SCED dispatch shall be within the Pmax and Pmin declared by the

participating generator.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 22

 3 SCED DEVELOPMENT & OPERATION IN 2020

3.1 Implementation of RTM

Real-Time Market (RTM) in India was operationalized from 1st June 2020

onwards. RTM provides another avenue to generators to sell their un-requisitioned

surplus (URS) in the market after all constituents have exhausted the opportunity to

revise their requisitions. Trades in RTM take place an hour before dispatch at half-
hourly intervals. To facilitate quick and automatic incorporation of RTM trades in

schedules, an in-house software application to provide an interface to power

exchanges and WBES and for verification of margins and facilitating RTM clearing was

developed at NLDC. The secure API developed for SCED implementation was
leveraged for data exchange between and WBES and RTM applications.

With RTM in operation, RRAS and SCED dispatch is taking place after RTM and
the URS available post-RTM and RRAS is used for SCED optimization. For faster

dissemination of schedule data to for bidding decision in RTM, a secure API has been
developed and shared with generators and constituents. This API also enables faster

dissemination of injection schedules after incorporation of RRAS and SCED quantum.

3.2 New Scheduling Timelines

With the implementation of RTM, new scheduling timelines came into vogue
with the gate closure for DC and requisition revisions getting shifted from 4th time

block to 7th/8th time block before dispatch block. The intervening blocks were
earmarked for RTM auction, clearing and dissemination of results. SCED continued to

run 20 minutes before dispatch, i.e. in T-2 block. T-3 block was earmarked for dispatch

and incorporation of RRAS. With operationalization of algorithmic RRAS dispatch

algorithm on 19th September 2020, the requirement of one separate block for

dispatch and incorporation of RRAS was eliminated, as RRAS and SCED outputs started
getting calculated together. Therefore, SCED run was shifted to 35 minutes before
dispatch (T-3 block), giving generators greater time to prepare for any ramp events.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 23

 COMPOSITE TIMELINE FOR DESPATCH OF SCED AND RRAS

TIME → T-90 T-75 T-60 T-45 T-30 T-15 T (start of despatch period) T+15

X-3
X-6 X-2 X-1
X-5 X-4 Schedule incorporation and IR synchronization; X
BLOCK → Right to recall SCED & RRAS Preparation
RTM Auction Window RTM Clearing Window RRAS punching gate closure; (delivery block)
ends Incorporation for dispatch
RRAS + SCED solution

TIME → T-73 T-60 to T-55 T-55 to T-52 T-52 to T-45 T-43 T-39 T-35 T-33 T-28

SCED Pilot - Detailed Feedback Report – Mar 2021

Final clearing of Fetching all
Running Fetching &
Provisional RTM and ISGS
Checking First run of Second RRAS+SCED incorporating
clearing of trades pushed schedules
WBES run provisional schedule to run of optimization RRAS and
All revisions of RTM and to NLDC RTM and
to finalize trades for incorporate schedule and SCED
DC, requisitions, provisional DB & onwards constraints Generate as per net
pre-RTM RTM auction for congestion RTM trades for determination UP/DOWN
ACTIVITY → URS punched on trades for to WBES DB and IR flow schedule (inclusive of
schedule blocks X and X+1 and and firm syncing of RRAS and from NLDC

September 2020)
RLDC WBES regional (for regional schedules RRAS and SCED)
for blocks transmitting up inter- SCED and
portal entities entities); from RLDCs
X and X+1 margins to State/ISGS regional UP/DOWN publication
pushed to and from PX to WBES. Gate
PX schedules schedule considering all of final ISGS

Gate closure for X & X+1

NLDC RTM DB SLDCs (for closure for
constraints schedules
state entities) RRAS input.
(implemented since 19th September 2020) is shown at Figure 1.

AGENCIES Regional entities, IEX, PXI, NLDC

ISGS, States, RLDCs NLDC (RTM RLDCs RLDCs NLDC (WBES,
INVOLVED state entities, IEX, PXI NLDC (RTM (RRAS+SCED RLDCs (WBES) ISGS ISGS
RLDCs (WBES) (WBES) app) (WBES) (WBES) RRAS app)
→ IEX, PXI app, WBES) engine, WBES)

* All time offsets in minutes from T (start of despatch period)

Figure 1 Composite Timeline for dispatch of SCED and RRAS (w.e.f. 19th
The diagram indicating detailed timelines for RTM, RRAS and SCED incorporation

Page 24
 3.3 Higher ramp rates in thermal ISGS
Regulation 30(2) of the CERC (Terms and Conditions of Tariff) regulations stated

that thermal generating stations would need to provide minimum ramp rate of
1%/min from 1.4.2020 onwards, failing which, their rate of return on equity would be

reduced by 0.25%. In compliance with the regulations, thermal ISGS have started
declaring ramp rates of 1%/min in WBES and these ramp rates are being incorporated

in SCED dispatch. The change in ramp rates from April 2020 is seen in the envelope of
blockwise ramp in post-SCED schedule of a sample station (Figure 2).

Block-wise ramp in post-SCED schedule - Talcher-II

400

300

200

100
Ramp (MW/block)

-100

-200

-300

-400

Figure 2 Block-wise ramp in post-SCED schedule of Talcher-II (Jan-Dec 2020)

The regulations also directed NLDC to formulate detailed guidelines for assessment
of ramping capability. Accordingly, guidelines were formulated and published in Feb

2020 after stakeholder consultation7. After the operationalization of these provisions,

many generating stations expressed difficulty in meeting the scheduled ramp rates

7Final detailed guidelines for assessment of ramping capability of thermal Interstate generating stations (ISGS)
https://posoco.in/download/final-detailed-guidelines-for-assessment-of-ramping-capability-of-
isgs/?wpdmdl=27458

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 25

during initial blocks of ramping. These challenges arise from the discretization of
continuous generation and ramping curves into 15-minute average MW for the

purpose of scheduling and metering. This aspect was recognized in the joint POSOCO-
NREL paper on Ramping capability in the Indian power system8. In order to address

these practical difficulties, the methodology of handling ramp of thermal generators

in the scheduling process has been formulated in consultation with NTPC. This

modified approach would be implemented in WBES and in SCED and has been
described in Annexure-2.

3.4 Impact of COVID-19 pandemic

Pan-India lockdown was imposed by Govt. of India for containment of COVID-19

pandemic from 25th March 2020 onwards, which was relaxed gradually after May 2020.
All-India Electricity demand reduced by up to 25-30% due to shutting down of a large

chunk of industrial, commercial and traction loads. The demand gradually recovered

as lockdown was progressively relaxed and economic activities restarted across the
country. With reduction in demand, a significant generation capacity went under

reserve shutdown. This aspect is shown in Figure 3 where the reduction in DC on-bar
and net schedule of SCED generators towards end of March 2020 can be seen.

8 Ramping Up the Ramping Capability – India’s Power System Transition, National Renewable Energy
Laboratory, September 2020
https://www.nrel.gov/docs/fy20osti/77639.pdf

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 26

 Average net schedule of SCED generators (Jan-Dec 2020)
All India Energy Met DC On-bar Average Net Schedule
50000 4500
Daily Avg DC On-bar & net schedule under SCED (MW)

45000 4000

Daily All India Energy Met (MU)

40000 3500

35000 3000

30000 2500

25000 2000
Mar-20

May-20

Nov-20
Apr-20

Aug-20

Sep-20

Oct-20

Dec-20
Jan-20

Feb-20

Jun-20

Jul-20

Figure 3 Average Net Schedule of SCED generators & All India Energy met

Daily distribution of SCED System Marginal Price (Jan-Dec 2020)

450 95th percentile 5th percentile Median

400

350
SMP (paise/kWh)

300

250

200

150

100
Apr-2020

Nov-2020
Mar-2020

May-2020

Dec-2020
Jan-2020

Feb-2020

Jun-2020

Aug-2020

Sep-2020

Oct-2020
Jul-2020

Figure 4 Distribution of SCED SMP

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 27

Figure 4 shows the distribution of SCED System Marginal Price over 2020 in terms of
its 5th percentile, median and 95th percentile on each day. There is a sharp dip towards

late-March 2020 due to demand reduction.

3.5 New features in SCED

SCED application has been continuously improved over time with addition of new
features. Some of these are:

3.5.1 Modular Integration with RRAS

RRAS was operationalized in April 2016 through an in-house developed software that

dispatched RRAS through a bucket-filling approach, which was then sent to RLDCs
through FTP/email, where it was incorporated in WBES. Certain issues had been

observed over time that were hampering effective dispatch of RRAS with this approach,
namely:

 Recall by original beneficiaries resulting in non-scheduling of dispatched RRAS,

as RRAS dispatch window straddled beyond gate closure boundary.

 Requirement of manual inputs and telephonic confirmation in dispatch and

incorporation of RRAS in net schedules of generators within tight scheduling

timelines

 Data exchange reliability issues due to usage of older technologies like FTP and
csv files

Post-RTM implementation from 1st June 2020, the problems in effective dispatch of
RRAS became more acute as RRAS could now be dispatched after clearing of RTM only

and each instruction could not exceed 2 blocks at a time. A window of 10 minutes was
available to NLDC shift engineers for dispatching RRAS, intimating RLDCs and its

incorporation in schedules and this process had to be repeated every 30 minutes.

In view of the above mentioned issues, it became necessary to redevelop the RRAS

dispatch software implementation. As significant experience had been gained with

optimization-based formulations and the data exchange requirements are similar to

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 28

SCED implementation, the new software was built with an optimization-centric base
and harnessing the SCED technology infrastructure. The new RRAS software recognizes

the fact that while a look ahead reserve position is important for the real time operator,
RRAS dispatch shall be only for the blocks for which RTM period is over.

A modularized approach has been introduced for handling SCED and RRAS in a
common framework. The optimization process works on a common generator stack,

which is the superset of SCED and RRAS generators. Despite the common platform,
RRAS and SCED are governed in accordance with their existing regulatory structures.

New RRAS implementation seeks to emulate the process prevailing earlier with the

RRAS dispatched forming the part of input schedule for SCED. As the regulatory and
commercial treatment of RRAS and SCED are different, both continue to be computed

and shown in schedules separately.

User
Interface for
RRAS

Extra RRAS data Extra RRAS data

New RRAS
Core Engine Application Scheduling
RRAS + SCED
RRAS data Program software
DB
SCED Interface (API) WBES
Core Engine

Patch for accounts

Mimic insertion of
RRAS dispatch into
Existing user WBES schedule
interface using GAMS “Loop”

Figure 5 Architectural overview of RRAS implementation

A high level view of the integration of SCED and RRAS applications is shown in
Figure 5. RRAS output gets incorporated automatically in net schedules of generators,

without need for manual action or telephonic confirmation. The communication takes
place through API which improves reliability and eliminates the need for CSV file
exchange through FTP/email.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 29

 As RRAS dispatch takes place after gate closure for constituents and RTM clearing,
the certainty of obtaining desired RRAS quantum is much higher than with the earlier

implementation. Reserve accumulation to ensure max utilization of spinning reserves

is ensured by coding discrete-time integration of reserves.

Although determined separately, both SCED and RRAS outputs are written together to
WBES, resulting in availability of an additional block to generators to prepare for

ramping (This can be seen in Figure 1). Features of the new RRAS software have been
described in Annexure-3.

3.5.2 Incorporation of Export ATC

Initially, SCED application incorporated the import ATC and inter-regional schedules
towards drawing regions. With growth in renewables, the inter-regional flow patterns

are undergoing a seasonal and diurnal shift. In view of this shift, the SCED software has
been modified to incorporate import as well as export ATC constraints on inter-

regional schedules for all regions. The user ATC feature has also been modified to give
the system operators the ability to restrict export flows from regions also in case of

congestion. Further, import of TTC/ATC data in WBES has been centralized at NLDC to
facilitate faster revisions in case of contingencies in the grid and also to ensure that

the right margins go to the RTM.

3.5.3 Soft landing on exclusion

Generators have to be excluded at times from SCED due to certain requirements

like testing or integration activities, system requirements etc. Earlier, whenever any

station was excluded from SCED, the post-SCED schedule was made equal to the pre-
SCED schedule from the next time block and stations were asked to return to pre-SCED

schedule following their respective ramp rates. This also necessitated post-facto
correction of SCED schedules to incorporate ramp. To ease this aspect, a soft-landing

feature has been added to SCED software, wherein whenever any station is excluded
from SCED, its post-SCED schedule returns to the original schedule following the

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 30

declared ramp rate. This provides generators a clear guideline in real-time operation
and obviates the need for post-facto correction of schedules for this aspect.

Figure 6 shows a sample exclusion event with the soft-landing feature.

Station
excluded
from SCED

Figure 6 Soft landing on exclusion

3.5.4 Scheduled and Emergency Stop

The ability to stop SCED optimization had been provided in the SCED software
dashboard in case of any technical/communication issues impairing the ability of SCED

and WBES implementation to deliver optimized schedules to generators. When SCED

was stopped, all stations were excluded from SCED and their schedules were made

equal to original schedules from the next block onwards. Generators were asked to
return to their original schedules with their respective ramp rates. The SCED algorithm

continues to work with all generators excluded.

The need for stopping SCED as a planned activity in certain cases was recognized, such

as on 5th April 2020 during the 9 pm 9 minute event. Therefore, the SCED stop
functionality was split into two types depending on the nature of event requiring SCED

to be stopped – Scheduled Stop and Emergency Stop.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 31

 Figure 7 Scheduled and Emergency Stop in SCED

Scheduled Stop functionality is intended to be used when SCED has to be stopped in

a planned manner and the ability to deliver optimized schedules to generators through

SCED and WBES is not affected. This is analogous to excluding all generators from
SCED with soft-landing. All generators are brought back to their original schedules

following their ramp rates and the SCED equality constraint is satisfied throughout this
process.

Emergency Stop functionality is intended to be used when the ability of the SCED and
WBES implementation chain to deliver optimized schedules to all stations is impaired

due to communication issues between one or more RLDCs and NLDC, issues affecting
WBES at any RLDC, or technical issues at NLDC. In this case, as the ability to determine

and/or deliver optimized schedules is not assured, soft-landing algorithm cannot be

utilized. Optimized schedules are made equal to original schedules and SCED

component is made zero from the next block onwards. As RRAS also utilizes the same
IT implementation for data input, dispatch and dissemination, with Emergency Stop

activated, RRAS output is also made zero from the next block onwards. Another layer
of software interlock flags is put in place to ensure that any inadvertent incorrect

SCED/RRAS value does not get incorporated in schedules. Not all technical issues may
hamper the delivery capability in all regions, as these issues may remain localized to

one region only. Therefore, Emergency Stop is modelled on regional basis – emergency
stop can be applied in one region while SCED and RRAS for generators of other regions

continue to operate. Figure 7 shows the region selection pop-up when Emergency

Stop is used.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 32

 3.5.5 Evacuation Limit

Due to contingencies in the power system, in order to maintain reliability, the

generation from one or a group of generators may need to be restricted during certain
periods. On certain other instances, the generation from some generators might need

to be kept above a certain level. Earlier, in such cases, these generators were excluded
from SCED and their schedules were being curtailed through WBES. For quicker and

more effective control of generation in such scenarios, the evacuation limit feature has
been added to SCED. Real-time shift engineers can activate this feature for any set of

generators and specify station-wise maximum/minimum evacuation limit, which is

then ensured by SCED.

The SCED formulation, including all formulae and equations, has been described in

detail in Annexure-4.

3.6 Reduction in Costs

During 2020 (Jan-Dec), the total reduction in variable cost of generation due to SCED
was ₹ 752 Crores. From the start of the SCED pilot in April 2019 up to January 2021,

the cumulative cost reduction is ₹ 1624 crores. Table 1 shows the generator-wise
energy increment and decrement due to SCED during the pilot period along with the

money flowing into/out of the SCED pool account on account of the

increment/decrement. Figure 8 shows cost reduction due to SCED on daily basis during

the period January 2020 to January 2021.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 33

 Daily cost reduction due to SCED (Jan 2020-Jan 2021)
0
All India 1% ramp RTM Revamped
Lockdown RRAS

-100
Cost Reduction (₹ Lakh/day)

-200
All-time
high
demand
-300

-400

-500

-600

Figure 8 Daily cost reduction due to SCED (Jan 2020-Jan 2021)

It is observed that the daily SCED savings have reduced towards the second half of

2020. Part of this reduction could be due to reduced perturbation (increment and
decrement of generation). Figure 9 shows the trend of daily average perturbation in
SCED which shows a decreasing trend. Figure 13 shows the perturbation caused by
SCED in Dec 2020 vs Dec 2019. It can be seen that the perturbation has reduced

significantly in Dec 2020 as compared to the previous year. The corresponding cost
reduction during Dec 2019 and Dec 2020 was ₹ 68 Crore and ₹ 28 Crore respectively.

The reduction in diversity of Variable Charges in the SCED merit order stack (discussed
in Section 3.8) could be another contributor to reduced savings. This aspect can be

seen in Figure 10 as well, where it is seen that in second half of 2020, savings are lower
for the same MW quantum of perturbation9.

9 SCED Savings = Σ(VC*perturbation), for all plants

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 34

 Average Perturbation in SCED (Apr 2019 onwards)
3000

2500

2000
Perturbation (MW)

1500

1000

500

Figure 9 Daily Average Perturbation in SCED

Average Perturbation vs Cost Reduction

800 2019 H1 2020 H2 2020

700

600
Cost Reduction (Rs. Lakh/day)

500

400

300

200

100

0
0 500 1000 1500 2000 2500
Perturbation (MW)

Figure 10 Average SCED Perturbation vs Daily Cost Reduction

As RTM provides an avenue to generators to sell their unrequisitioned power after gate

closure for constituents, it is generally expected that RTM sale by SCED generators
would diminish the margin available for SCED to perturb generation and reduce
variable cost of generation. On average, the daily quantum of sale in RTM by SCED

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 35

generators was 2.5 MU, which is quite miniscule (Figure 11). To get a rough idea of this
impact in quantitative terms, a sample study was carried out for the period 1 st to 15th

December 2020. SCED operation was simulated on a test bench for this period
considering ISGS schedules with and without RTM trades. On overall basis, the study

results show that the SCED savings reduced by 3% (from ₹1327 Lakhs to ₹1287 lakhs)
with incorporation of RTM. Figure 12 shows blockwise cost reduction under both

simulations with RTM trades by thermal ISGS.

Daily RTM Sale Volume by SCED Generators

10

6
Daily Volume (MU)

0
01-Jun-20 01-Jul-20 01-Aug-20 01-Sep-20 01-Oct-20 01-Nov-20 01-Dec-20 01-Jan-21

Figure 11 Daily Sale Volume in RTM by SCED generators

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 36

 Impact of RTM on SCED cost reduction
RTM Trades Cost Reduction-without RTM Cost Reduction-with RTM
10 1600
Negative sign implies reduction of cost

5 1400

0
1200

-5
Cost Reduction (₹ Lakh/hr)

RTM sale by ISGS (MW)

1000

-10
800
-15

600
-20

400
-25

-30 200

-35 0

Figure 12 Impact of RTM Trades on SCED Cost Reduction – study results

Duration curve of perturbation by SCED: Dec 2019 vs Dec 2020

4000 Dec-20 Dec-19

3500

3000
Perturbation (MW)

2500

2000

1500

1000

500

0
0 10 20 30 40 50 60 70 80 90 100
Percentage of time

Figure 13 Perturbation by SCED: Dec 2019 vs Dec 2020

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 37

 Table 1 Generator wise SCED statement (Apr 2019 – Jan 2021)

Charges To
Charges To be Net
be Paid to
Refunded by Charges
Increment Decrement SCED
SCED Payable
due to due to Generator
Generator to (+) /
SI. SCED SCED SCED from
Region National Pool Receivable
No. Generator scheduled scheduled National
(SCED) (-)
to VSCED to VSCED Pool (SCED)
(in ₹ Crore) (in ₹ Crore)
(MU) (A) (MU) (B) (in ₹ Crore)
(D) = (B) x (E)* = (C) –
(C) = (A) x
V.C. (D)
V.C.
1 BARH ER 544 1,059 135 281 -146
2 BRBCL ER 490 428 110 101 9
3 FSTPP-I&II ER 936 635 240 165 75
4 FSTPP-III ER 243 84 62 21 41
5 KHSTPP-I ER 682 177 154 40 114
6 KHSTPP-II ER 1,142 153 244 33 211
7 MTPS-II ER 281 213 76 59 18
8 TSTPS-I ER 791 214 160 44 117
9 MPL ER 515 696 137 188 -51
10 NPGC ER 246 212 51 45 6
11 Darlipali STPS ER 71 0 8 0 8
Ramagundam
12 SR 807 1,465 207 382 -175
STPS (U1-6)
Ramagundam
13 SR 206 289 52 74 -22
STPS (U7)
Simhadri STPS-
14 SR 207 760 65 239 -174
II
Talcher STPS
15 SR 703 581 141 119 22
Stage II
16 Vallur TPS SR 50 1,288 17 474 -457
NLC TPS-II
17 SR 48 68 13 18 -6
Stage I
NLC TPS-II
18 SR 63 89 17 24 -7
Stage II
NLC TPS-I
19 SR 27 61 7 15 -9
Expansion
NLC TPS-II
20 SR 8 13 2 3 -1
Expansion
21 NTPL SR 440 666 128 197 -69
Simhadri STPS -
22 SR 146 1,133 46 363 -317
I
23 Kudgi - I SR 96 713 30 274 -244
24 CGPL WR 760 101 143 19 124
25 KSTPS I&II WR 225 135 31 20 11
26 KSTPS7 WR 17 14 2 2 0
27 MOUDA WR 247 676 75 205 -130
28 MOUDA_II WR 249 752 75 229 -154
29 NSPCL WR 87 943 24 298 -274

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 38

 Charges To
Charges To be Net
be Paid to
Refunded by Charges
Increment Decrement SCED
SCED Payable
due to due to Generator
Generator to (+) /
SI. SCED SCED SCED from
Region National Pool Receivable
No. Generator scheduled scheduled National
(SCED) (-)
to VSCED to VSCED Pool (SCED)
(in ₹ Crore) (in ₹ Crore)
(MU) (A) (MU) (B) (in ₹ Crore)
(D) = (B) x (E)* = (C) –
(C) = (A) x
V.C. (D)
V.C.
30 SASAN WR 630 11 84 1 82
31 SIPAT I WR 228 110 33 17 16
32 SIPAT II WR 161 74 24 11 12
33 SOLAPUR WR 62 331 19 101 -83
34 VSTPS I WR 633 188 109 32 77
35 VSTPS II WR 265 163 44 27 17
36 VSTPS III WR 369 97 60 16 45
37 VSTPS IV WR 351 78 57 13 44
38 VSTPS V WR 284 58 48 10 38
39 GADARAWARA WR 371 138 97 40 57
40 LARA WR 378 191 86 45 41
41 Khargone WR 218 158 60 43 16
42 JHABUA WR 27 46 7 12 -5
43 DADRI TPS NR 12 467 4 185 -180
44 DADRI-II TPS NR 32 918 11 334 -323
IGSTPS-
45 NR 77 661 28 231 -203
JHAJJAR
46 RIHAND STPS NR 781 9 108 1 106
47 RIHAND-II STPS NR 698 10 96 2 94
RIHAND-III
48 NR 684 7 93 1 92
STPS
SINGRAULI
49 NR 1,478 16 206 2 204
STPS
50 TANDA-II STPS NR 454 126 112 32 80
UNCHAHAR-I
51 NR 92 311 29 102 -73
TPS
UNCHAHAR-II
52 NR 64 323 21 107 -86
TPS
UNCHAHAR-III
53 NR 59 150 19 50 -31
TPS
UNCHAHAR-IV
54 NR 255 234 79 73 6
TPS
55 BONGAIGAON NER 187 735 61 248 -187
All India Total 19,175 19,230 4,043 5,666 -1,624

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 39

 3.7 Ease of generator operations
As discussed in earlier feedback report, SCED has brought more certainty in the

dispatch pattern of generators. There is an average decrease of 19% in the cumulative

MW change in schedules and average decrease of 44% in the number of instructions

(counting any change in the schedule of the generator by the RLDC as one instruction)
to all the generators during the period Jan-Dec 2020. It has resulted in increased plant

load factor in lower variable cost power stations & vice versa. Therefore, pilot on SCED
has facilitated the ease of generators’ operations by requiring change in generation

schedule less often.

The generator-wise Pre-SCED and Post SCED Change in Number of Instructions

and cumulative MW change in schedules (Jan-Dec 2020) is tabulated in Table 2.

Table 2 Generator-wise change in number of instructions and Delta MW

VC Instruction Count % change Cumulative Delta MW %

Generator Post Pre in Post Pre change
(p/kWh) instructions in MW
SCED SCED SCED SCED
Sasan 132.1 1542 3733 -58 206402 429687 -51
Darlipali 133.8 764 2582 -70 30525 87255 -65
Singrauli TPS 139.6 2920 10689 -72 252665 789122 -67
Korba-
141 2689 2701 0 61507 57821 6
Stage-3
Korba STPS
Stage 1 and 143.9 2075 4619 -55 272083 307703 -11
2
Rihand 3 147.4 1060 8788 -87 94762 415388 -77
Sipat STPS
147.4 2073 4463 -53 245495 269873 -9
Stage 1
Rihand 2 149.3 1231 9383 -86 119708 478719 -74
Rihand 1 149.6 1808 10407 -82 157312 510911 -69
Sipat STPS
151.9 2079 4451 -53 173038 188780 -8
Stage 2
Vindhyachal
167.3 3118 6593 -52 239641 319564 -25
STPS Stage 4
Vindhyachal
169.3 3129 6827 -54 246678 316977 -22
STPS Stage 3
Vindhyachal
170.7 3755 6354 -40 300318 315621 -4
STPS Stage 2
Vindhyachal
172.8 3402 7473 -54 166436 227212 -26
STPS Stage 5
Vindhyachal
177.3 6246 12836 -51 403408 530646 -23
STPS Stage 1
CGPL
189.01 11408 11410 0 1169307 1173221 0
Mundra

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 40

 VC Instruction Count % change Cumulative Delta MW %
Generator Post Pre in Post Pre change
(p/kWh) instructions in MW
SCED SCED SCED SCED
NPGC 202.5 5215 7769 -32 275243 337140 -18
Talcher
Stage 2- 203.4 6690 8168 -18 941938 1048877 -10
NTPC
Kahalgaon
207.4 5929 15774 -62 583921 867853 -32
Stage 2
Talcher 214.3 6571 12510 -47 515111 711548 -27
Kahalgaon
218.6 6720 16191 -58 477694 613424 -22
Stage 1
Nabinagar 227.4 7403 23591 -68 408202 418812 -2
Tanda-2 239.7 7413 9935 -25 475355 507907 -6
FARAKKA-
248.7 7477 7623 -1 358830 270315 32
STAGE-3
Barh 252 8324 17141 -51 817572 892499 -8
LARA 252 6541 10944 -40 461597 569670 -18
Farakka
Stage 1 and 253.4 8144 12859 -36 986877 997517 -1
2
Ramagundam
258.1 4877 5231 -6 258732 239222 8
TPS Stage 3
Ramagundam
TPS Stage 261.9 10946 15570 -29 1549383 1520905 1
1&2
MTPS Stg 2 265.5 8165 8543 -4 283818 252465 12
MPL 267.9 9446 13779 -31 784362 760992 3
Jhabua 270.5 2055 1758 16 110881 92936 19
NTPL 279.7 8619 9846 -12 755328 672323 12
NTPC-
292 7653 9321 -17 555721 567136 -2
KHARGONE
Gadarwara-
298 5907 7947 -25 366735 397663 -7
1
Mauda 2 311.7 8151 10790 -24 541568 570010 -4
Mauda 314.7 7544 10765 -29 448521 447374 0
NSPCL 317.8 8747 12486 -29 331743 345477 -3
UNCHAHAR
328.8 5538 6491 -14 287736 295915 -2
4
NTPC
329.7 4092 5627 -27 308482 309498 0
SOLAPUR
Bongaigaon 331.9 5422 24961 -78 273649 514260 -46
Simhadri
NTPC Stage 334.5 7082 9068 -21 580730 679315 -14
2
Simhadri -
NTPC Stage - 338.8 6515 6766 -3 486669 593238 -17
1
Unchahar TPS
348.9 4528 6822 -33 164705 198796 -17
1
Unchahar TPS
348.9 4525 7539 -39 104753 127724 -17
3

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 41

 VC Instruction Count % change Cumulative Delta MW %
Generator Post Pre in Post Pre change
(p/kWh) instructions in MW
SCED SCED SCED SCED
Unchahar TPS
352 4160 7579 -45 161870 227029 -28
2
Indira
Gandhi TPS 370.1 3807 5524 -31 287861 368594 -21
(Jhajjar)
Vellur NTECL 380.1 4423 7985 -44 417456 649636 -35
Kudgi-1 380.5 3423 4783 -28 383510 549861 -30
Dadri Stage
386.1 3017 14889 -79 204610 629922 -67
2
Dadri Stage
413.7 1016 3153 -67 35180 69381 -49
1
All India 265384 473037 -43.9 20125628 24733734 -18.6

3.8 Variations in Variable Cost

The variable charges declared by the participating generators in SCED undergo change
on month-on-month basis. Table 3 depicts the temporal variation of variable charges

across India in the form of a heat map. Further, Table 4 depicts the change
(increase/decrease) in variable charges of SCED generators from previous month. Table

5 indicates the difference of VC between successive stations in the SCED merit order
stack. It is the difference in VC between the high merit and low merit stations which

generates the savings in SCED. It is readily apparent from Table 3 and Figure 14 that
over time, the variable charges of costlier stations have come down. This may be on
account of optimization in coal allocations and coal movement by the generating

stations. The maximum VC, which went up to 443 paise/kWh in Oct 2019 has come
down to 317 paise/kWh in Jan 2021. The standard deviation of VC within the SCED

stack has reduced from 87.6 paise/kWh in Jun 2019 to 58.6 paise/kWh in Jan 2021.

It follows that with reduced variation in variable charges, the potential for cost

reduction also comes down, which may explain the trend towards lower cost reduction
seen in later half of 2020.

Figure 15 shows the trend of daily average cost of generation since the
commencement of SCED pilot. It is observed that post April 2020, there was decrease
in the average cost, due to the impact of costlier machines going off-bar during the

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 42

 Month-wise variable cost of SCED generators (paise/kWh)
2019 2020 2021
Sl No Plant name Region 16 Mar 16 Apr to 16 May 16 Jun 16 Jul to 16 Aug 16 Sep 16 Oct to 16 Nov 16 Dec 16 Jan 16 Feb to 16 Mar 16 Apr to 16 May 16 Jun 16 Jul to 16 Aug 16 Sep 16 Oct to 16 Nov 16 Dec 16 Jan
to 15 Apr 15 May to 15 Jun to 15 Jul 15 Aug to 15 Sep to 15 Oct 15 Nov to 15 Dec to 15 Jan to 15 Feb 15 Mar to 15 Apr 15 May to 15 Jun to 15 Jul 15 Aug to 15 Sep to 15 Oct 15 Nov to 15 Dec to 15 Jan to 15 Feb
1 Darlipali STPS ER 134 119 109 104 103 103 102 103 105 110 104
2 Sasan WR 132 132 132 132 133 132 133 133 132 132 132 132 132 132 132 132 136 136 135 135 135 135 135
3 SIPAT-I WR 118 120 128 127 132 141 140 151 176 147 144 150 145 136 156 149 140 136 145 150 139 129 135
4 SIPAT-II WR 122 123 132 131 137 146 144 156 181 152 149 155 150 140 160 153 144 141 149 155 143 133 139
5 Rihand-III NR 133 135 129 127 142 133 130 130 133 147 135 136 130 132 138 142 142 143 146 143 137 137 140
6 Korba(III) WR 122 126 125 130 129 125 133 144 136 141 130 130 136 142 139 139 137 131 127 130 152 145 141
7 Rihand-II NR 132 133 131 129 144 135 132 131 135 149 137 138 132 134 140 144 144 145 147 145 139 139 142
8 Rihand -I NR 132 134 131 129 144 135 132 131 135 150 137 138 132 134 140 144 144 145 148 146 139 139 142
9 Singrauli NR 139 138 137 144 146 141 134 136 138 140 141 138 136 135 138 143 143 145 140 141 144 143 143
10 Korba(I & II) WR 124 128 128 133 132 128 136 147 138 144 132 133 138 144 142 142 139 134 129 132 155 148 144
11 Vindhyachal-IV WR 149 153 153 154 157 181 170 164 157 167 161 166 166 166 167 157 161 159 154 154 145 151 159
12 Vindhyachal-III WR 149 153 155 156 159 183 172 166 160 169 164 168 168 169 169 159 163 161 156 155 147 153 161
13 Vindhyachal-II WR 149 153 156 158 160 184 173 168 161 171 165 169 169 170 170 160 164 162 157 157 149 154 162
14 Vindhyachal-V WR 150 153 155 156 162 186 175 170 163 173 163 171 171 172 172 162 166 164 159 159 151 156 165
15 Vindhyachal-I WR 159 163 162 163 166 191 180 174 167 177 171 175 176 176 177 166 171 168 163 163 154 160 169
16 Coastal Gujarat PL WR 180 181 177 197 195 194 189 194 192 189 190 185 192 194 197 199 198 197 197 196 193 189 187
17 Talcher-II SR 168 174 190 208 191 182 216 239 213 203 195 206 204 208 200 189 191 201 204 210 211 206 188
18 Talcher-I ER 167 174 191 210 193 183 218 241 214 205 197 208 200 209 201 191 193 203 206 212 213 207 189
19 LARA WR 233 233 230 252 271 243 239 247 236 223 217 208 199 186 191 202 215
20 Kahalgaon-II ER 215 226 209 213 223 208 210 211 207 216 215 218 217 215 214 214 209 200 204 217 228 221 221
21 NPGC ER 231 227 206 203 213 205 208 212 202 192 190 201 203 221 224 223 228
22 Kahalgaon-I ER 225 237 221 224 236 220 221 222 219 228 226 230 229 226 226 225 221 201 215 229 240 234 234
23 BRBCL ER 195 191 216 219 216 222 223 219 229 227 232 235 244 243 239 239 249 250 250 231 229 227 236
24 Ramagundam- III SR 257 256 264 251 252 266 266 273 258 258 259 259 257 256 254 254 243 246 246 233 247 244 237
lockdown-induced low-demand period.

25 Ramagundam- I & II SR 262 262 269 255 256 270 271 277 263 262 263 263 260 260 258 258 247 250 249 236 250 248 241
26 MPL ER 272 273 274 269 269 267 268 266 261 263 264 266 266 275 277 271 261 275 250 248
27 GADARAWARA WR 377 377 373 373 338 326 330 298 293 311 304 304 294 254 249 250 247 247 253 255 254

SCED Pilot - Detailed Feedback Report – Mar 2021

28 NLC- I Exp SR 236 248 242 243 241 242 242 243 255 255 255 255 255 254 254 254 255 256 255 255 249 253 254
29 Barh ER 221 220 218 226 234 237 247 248 245 252 285 285 301 321 303 274 260 266 259 269 255 269 255
30 Tanda TPS Stage-II NR 234 240 239 243 238 246 260 255 252 253 255 254 256 259 257
31 NTPC SAIL PCL WR 251 251 305 357 326 329 380 429 351 318 333 345 311 302 259 258 267 259 252 259 260 262 259
32 Farakka-III ER 258 239 252 242 212 235 229 236 249 250 247 274 266 270 289 289 277 256 255 264 268 264 259
33 Farakka-I&II ER 256 237 256 246 216 240 233 241 253 255 252 280 271 275 295 294 282 261 260 269 273 269 264
34 JHABUA WR 276 265
35 NLC- II Exp SR 236 248 248 248 248 248 248 248 260 260 259 259 256 259 257 256 257 257 257 264 265 266 266
36 Unchahar-IV NR 285 283 303 301 312 322 324 328 322 329 329 325 325 335 343 316 306 283 295 293 285 273 267
37 Khargone WR 292 297 310 255 259 258 280 278 278 278 267
38 Simhadri- II SR 323 310 332 322 319 317 315 334 325 335 348 316 314 306 299 290 295 300 310 306 287 277 269
39 Mouda-I WR 277 288 290 312 316 328 335 332 326 315 311 309 314 312 286 286 289 276 294 293 279 270 270
40 Simhadri- I SR 322 310 336 326 323 320 319 338 329 339 352 320 318 310 303 293 298 303 314 310 290 280 273
41 Kudgi STPS SR 400 396 400 373 376 384 379 398 405 381 370 361 360 386 359 319 313 311 337 335 286 276 276
42 NLC- I SR 255 269 264 264 264 266 264 264 278 277 278 277 277 276 277 277 277 277 277 277 277 278 278
43 NLC- II SR 255 269 264 264 264 266 264 264 278 277 278 277 277 276 277 277 277 277 277 277 277 278 278
44 Vallur SR 372 379 381 381 407 391 368 378 376 380 403 369 360 365 356 353 353 334 324 314 308 291 279
45 MTPS-II ER 271 270 256 272 305 291 276 273 272 266 274 271 277 266 266 260 260 281 283 278 253 267 280
46 Unchahar-I NR 308 306 322 319 331 342 344 348 342 349 350 345 345 355 364 336 325 301 313 311 303 289 283
47 Unchahar-III NR 308 306 322 319 331 342 344 348 342 349 350 345 345 355 364 336 325 301 313 311 303 289 283
48 Unchahar-II NR 308 306 325 322 334 345 347 351 345 352 353 348 348 358 367 339 328 303 316 314 305 292 286
49 Dadri- I NR 357 378 392 385 401 401 421 443 439 414 421 419 384 388 385 378 370 374 365 324 314 293 290
50 SOLAPUR WR 385 389 409 435 395 352 317 331 337 330 328 310 320 313 329 285 282 284 294 285 287 290 291
51 NTPL SR 292 320 345 355 341 315 275 316 278 280 291 291 286 279 278 270 270 285 285 281 300 297 291
52 Mouda-II WR 283 282 281 318 327 322 338 340 331 312 307 303 311 314 288 289 289 285 295 294 288 291 292
53 Dadri -II NR 335 347 364 364 375 379 393 406 411 386 377 378 364 361 363 358 349 356 345 345 323 312 296
54 IGSTPS-JHAJJAR NR 359 353 350 369 374 364 376 364 368 370 359 355 355 357 357 349 349 333 334 333 329 328 316
55 Bongaigaon TPP NER 302 302 322 339 333 318 335 340 327 332 339 333 357 346 346 343 351 352 350 340 346 330 317
Minimum VC 118 120 125 127 129 125 130 130 132 132 130 130 130 119 109 104 103 103 102 103 105 110 104
Maximum VC 400 396 409 435 407 401 421 443 439 414 421 419 384 388 385 378 370 374 365 345 346 330 317
Max VC - Min VC 282 276 284 308 278 276 292 313 307 282 291 289 254 270 276 274 267 272 263 242 241 220 213
Mean VC 235 237 247 251 252 254 253 258 254 253 253 252 249 250 249 240 239 237 238 237 235 231 229
Table 3 Distribution of Variable Charges of SCED Generators (Apr-19 onwards)

Median VC 251 248 254 248 248 248 247 248 254 255 259 260 258 260 259 255 253 256 254 254 253 254 254

Page 43
Standard Deviation 81.4 81.6 87.5 87.6 86.6 83.5 82.9 86.9 82.8 78.3 82.1 78.0 77.1 78.8 76.9 71.8 70.3 68.0 70.1 67.4 64.3 61.8 58.6
 Month-on-month change in plant-wise VC of SCED generators (paise/kWh)
2019 2020 2021
Sl No Plant name Region 16 Apr to 16 May 16 Jun 16 Jul to 16 Aug 16 Sep 16 Oct to 16 Nov 16 Dec 16 Jan 16 Feb to 16 Mar 16 Apr to 16 May 16 Jun 16 Jul to 16 Aug 16 Sep 16 Oct to 16 Nov 16 Dec 16 Jan
15 May to 15 Jun to 15 Jul 15 Aug to 15 Sep to 15 Oct 15 Nov to 15 Dec to 15 Jan to 15 Feb 15 Mar to 15 Apr 15 May to 15 Jun to 15 Jul 15 Aug to 15 Sep to 15 Oct 15 Nov to 15 Dec to 15 Jan to 15 Feb
1 Darlipali STPS ER -15 -10 -5 -1 0 -1 0 3 5 -6
2 Sasan WR 0 0 0 1 0 1 0 0 0 0 0 0 0 0 0 3 0 0 0 0 0 0
3 SIPAT-I WR 2 8 -1 5 9 -2 12 25 -29 -3 6 -5 -10 20 -7 -9 -3 8 6 -12 -10 6
4 SIPAT-II WR 2 9 -1 6 9 -2 12 25 -29 -3 6 -5 -10 21 -7 -9 -3 9 6 -12 -10 6
5 Rihand-III NR 2 -6 -2 15 -9 -4 0 3 15 -13 1 -6 2 6 4 0 1 2 -2 -6 0 3
6 Korba(III) WR 4 -1 5 -1 -4 8 11 -9 6 -11 1 5 6 -2 0 -2 -6 -4 3 23 -7 -5
7 Rihand-II NR 2 -2 -2 15 -9 -3 -1 3 15 -13 1 -6 2 6 4 0 1 2 -2 -6 0 4
8 Rihand -I NR 2 -3 -2 15 -9 -3 -1 3 15 -13 1 -6 2 6 4 0 1 2 -2 -6 0 4
9 Singrauli NR 0 -2 8 2 -6 -7 2 2 2 1 -3 -2 -2 4 5 0 2 -5 1 4 -1 0
10 Korba(I & II) WR 4 0 5 -1 -4 8 11 -8 6 -12 1 5 6 -3 0 -3 -5 -6 3 23 -7 -4
11 Vindhyachal-IV WR 4 1 1 3 24 -11 -6 -7 10 -6 4 0 1 1 -11 4 -2 -5 0 -8 6 8
12 Vindhyachal-III WR 4 2 1 3 24 -11 -6 -7 10 -6 4 0 1 1 -11 4 -2 -5 0 -8 6 8
13 Vindhyachal-II WR 4 4 2 1 25 -11 -6 -7 10 -6 4 0 1 1 -11 5 -2 -5 0 -8 6 8
14 Vindhyachal-V WR 3 2 1 7 24 -11 -6 -7 10 -9 8 0 0 1 -11 4 -2 -5 0 -8 6 8
15 Vindhyachal-I WR 5 -1 1 3 25 -11 -6 -7 11 -7 5 0 1 1 -11 5 -2 -5 0 -9 6 8
16 Coastal Gujarat PL WR 1 -4 20 -1 -1 -5 5 -2 -3 1 -5 7 1 3 2 -2 -1 0 -1 -3 -4 -2
17 Talcher-II SR 6 16 19 -18 -9 35 23 -27 -9 -9 11 -2 5 -9 -10 2 10 3 6 1 -6 -18
18 Talcher-I ER 7 17 18 -17 -9 35 23 -27 -9 -9 11 -7 8 -8 -10 2 10 3 6 1 -6 -18
19 LARA WR 0 -3 22 19 -28 -4 8 -12 -12 -7 -9 -9 -13 5 12 12
20 Kahalgaon-II ER 11 -17 3 11 -15 2 1 -3 8 -1 3 -1 -2 0 -1 -4 -10 4 13 11 -6 0
21 NPGC ER -4 -21 -4 10 -8 4 4 -10 -10 -2 11 1 18 3 -1 5
22 Kahalgaon-I ER 11 -16 4 11 -16 2 1 -4 9 -1 4 -1 -2 0 -1 -5 -19 14 14 11 -7 0
23 BRBCL ER -4 26 3 -3 7 1 -5 11 -2 5 3 9 -1 -4 0 11 0 0 -19 -2 -1 9
24 Ramagundam- III SR -1 8 -13 0 14 0 8 -15 0 1 0 -3 -1 -2 0 -11 4 -1 -13 14 -3 -7
25 Ramagundam- I & II SR 0 7 -14 1 14 1 7 -15 -1 1 0 -3 0 -3 0 -11 3 -1 -13 14 -3 -7

SCED Pilot - Detailed Feedback Report – Mar 2021

26 MPL ER 1 1 -5 -1 -2 1 -2 -5 2 1 1 1 9 2 -6 -10 14 -25 -3
27 GADARAWARA WR 0 -4 0 -35 -11 4 -32 -6 18 -6 0 -10 -40 -5 1 -3 0 7 2 -1
28 NLC- I Exp SR 12 -6 1 -2 0 0 1 12 0 0 0 0 -1 0 0 1 1 0 0 -6 4 1
29 Barh ER -2 -2 9 7 3 10 2 -3 7 33 -1 16 20 -18 -30 -13 6 -7 10 -14 14 -14
30 Tanda TPS Stage-II NR 6 0 4 -5 8 14 -5 -3 1 2 -1 2 4 -2
31 NTPC SAIL PCL WR 0 54 52 -31 3 51 50 -79 -33 15 12 -34 -9 -43 -1 8 -8 -7 6 2 2 -3
32 Farakka-III ER -19 13 -9 -30 23 -6 7 12 2 -3 27 -9 4 20 0 -12 -21 -1 9 4 -4 -5
33 Farakka-I&II ER -19 19 -10 -30 24 -6 8 13 2 -3 28 -9 4 20 0 -12 -21 -1 9 4 -5 -5
34 JHABUA WR -11
35 NLC- II Exp SR 13 0 0 0 0 0 0 12 0 0 0 -4 3 -2 -1 0 0 0 7 1 1 0
36 Unchahar-IV NR -2 20 -2 11 10 2 4 -6 7 1 -4 0 10 8 -26 -10 -23 11 -2 -8 -13 -6
37 Khargone WR 5 13 -55 4 -1 22 -2 0 -1 -11
38 Simhadri- II SR -12 22 -11 -3 -2 -2 19 -9 10 13 -32 -2 -8 -7 -9 5 5 10 -4 -19 -10 -8
39 Mouda-I WR 11 3 22 4 12 8 -3 -6 -12 -4 -2 6 -2 -26 0 3 -12 18 -1 -14 -10 1
40 Simhadri- I SR -12 26 -10 -3 -3 -1 19 -9 10 14 -32 -2 -8 -7 -9 5 5 10 -4 -20 -10 -8
41 Kudgi STPS SR -4 4 -26 2 9 -6 19 7 -24 -11 -9 -1 27 -27 -40 -6 -2 26 -2 -49 -10 -1
42 NLC- I SR 14 -5 0 1 1 -1 0 14 -1 1 -1 0 0 0 0 -1 1 0 1 0 0 1
43 NLC- II SR 14 -5 0 1 1 -1 0 14 -1 1 -1 0 0 0 0 -1 1 0 1 0 0 1
44 Vallur SR 7 1 0 27 -17 -22 9 -2 4 23 -34 -9 5 -10 -2 0 -20 -10 -9 -6 -17 -12
45 MTPS-II ER 0 -15 17 33 -14 -15 -3 -1 -7 9 -4 7 -11 -1 -6 0 22 1 -5 -24 14 12
46 Unchahar-I NR -2 16 -3 12 11 2 4 -6 7 1 -4 0 10 8 -28 -11 -24 12 -2 -8 -14 -6
47 Unchahar-III NR -2 16 -3 12 11 2 4 -6 7 1 -4 0 10 8 -28 -11 -24 12 -2 -8 -14 -6
48 Unchahar-II NR -2 19 -3 12 11 2 4 -7 7 1 -4 0 11 9 -28 -11 -24 12 -2 -8 -14 -6
49 Dadri- I NR 21 15 -7 16 0 21 22 -4 -25 7 -2 -35 4 -4 -7 -8 5 -9 -42 -10 -20 -3
50 SOLAPUR WR 4 20 26 -40 -42 -36 14 6 -7 -2 -18 10 -7 16 -43 -4 2 10 -9 2 3 1
51 NTPL SR 28 25 10 -13 -27 -39 41 -38 2 11 0 -6 -7 -1 -7 0 14 0 -4 19 -3 -6
52 Mouda-II WR -1 -2 38 8 -4 16 2 -9 -19 -5 -4 8 3 -26 1 0 -4 11 -1 -6 3 1
53 Dadri -II NR 13 16 0 11 4 14 13 5 -25 -9 1 -14 -3 2 -6 -9 7 -11 0 -22 -12 -16
Table 4 Month-on-month change in Variable Charges (Apr-2019 onwards)

54 IGSTPS-JHAJJAR NR -6 -3 19 5 -9 12 -12 4 2 -11 -4 0 2 0 -8 0 -16 1 -2 -4 -1 -12

55 Bongaigaon TPP NER 0 20 16 -5 -15 16 6 -13 5 7 -6 25 -11 1 -4 8 1 -2 -10 6 -16 -13

Page 44
 Month-wise difference between VC of successive stations in SCED merit stack (paise/kWh)
2019 2020 2021
Rank
16 Mar 16 Apr to 16 May 16 Jun 16 Jul to 16 Aug 16 Sep 16 Oct to 16 Nov 16 Dec 16 Jan 16 Feb to 16 Mar 16 Apr to 16 May 16 Jun 16 Jul to 16 Aug 16 Sep 16 Oct to 16 Nov 16 Dec 16 Jan
Diff
to 15 Apr 15 May to 15 Jun to 15 Jul 15 Aug to 15 Sep to 15 Oct 15 Nov to 15 Dec to 15 Jan to 15 Feb 15 Mar to 15 Apr 15 May to 15 Jun to 15 Jul 15 Aug to 15 Sep to 15 Oct 15 Nov to 15 Dec to 15 Jan to 15 Feb
2-1 3 4 3 0 3 3 2 2 1 8 3 2 2 13 23 29 33 28 25 27 30 19 31
3-2 1 3 1 2 1 5 0 0 2 1 0 1 0 0 6 7 1 4 2 2 2 4 0
4-3 2 2 1 0 0 1 1 1 0 3 3 3 0 1 0 2 2 2 7 4 2 2 4
5-4 8 4 2 1 4 2 0 3 1 4 2 2 2 0 1 0 1 1 5 6 0 2 1
6-5 0 1 0 1 6 0 1 8 2 0 0 0 2 1 0 2 2 4 5 2 0 2 1
7-6 0 1 1 1 2 6 2 3 0 2 4 0 1 1 0 0 1 3 1 2 4 0 1
8-7 1 1 0 1 0 1 4 5 19 0 3 12 2 4 2 0 1 2 2 0 1 5 0
9-8 5 4 5 11 2 5 4 5 2 2 4 4 7 2 14 5 0 0 0 5 1 2 1
10-9 10 14 17 10 10 35 26 9 1 15 13 11 4 3 5 5 0 0 2 3 2 2 0
11-10 0 0 2 2 2 1 2 2 2 2 2 3 16 22 7 3 17 14 5 2 2 4 16
12-11 0 0 0 0 1 2 1 1 4 1 0 1 2 2 2 2 2 2 2 0 2 2 2
13-12 1 1 1 2 3 0 2 2 9 2 1 2 1 1 1 1 1 1 1 1 2 1 1
14-13 8 10 6 5 4 1 5 4 5 5 6 4 2 2 2 2 2 2 2 2 2 2 2
15-14 9 10 15 34 25 2 9 20 11 12 19 10 4 5 5 4 5 4 4 4 1 4 4
16-15 0 0 13 12 2 5 21 17 14 13 5 19 17 17 20 24 20 28 34 23 36 29 19
17-16 12 7 2 1 3 3 6 8 1 1 2 1 8 15 3 2 1 3 2 10 2 13 0
18-17 15 10 18 3 17 14 2 4 5 2 16 2 3 0 1 1 2 2 4 14 19 3 2
19-18 21 29 7 6 4 11 3 5 2 11 2 11 5 3 1 7 5 0 1 2 2 2 25
20-19 6 7 2 5 0 3 2 6 4 12 12 12 9 3 12 14 12 0 0 5 11 14 7
21-20 4 11 3 2 8 13 6 3 11 0 6 5 12 12 12 10 7 1 1 4 4 2 7
22-21 11 1 22 16 10 1 2 3 1 12 7 8 10 17 10 2 4 5 10 8 1 4 6
23-22 0 2 6 1 2 3 3 2 4 11 8 0 1 3 4 14 22 39 31 2 11 6 3
24-23 15 9 4 3 6 2 0 0 11 2 5 12 5 1 15 15 4 3 1 2 7 10 1
25-24 4 0 4 2 7 6 8 1 4 0 3 4 11 7 0 0 2 0 2 4 2 4 4

SCED Pilot - Detailed Feedback Report – Mar 2021

26-25 0 2 1 3 4 18 5 6 5 3 5 0 1 1 3 0 0 0 1 11 1 3 7
27-26 1 5 8 3 4 0 2 0 1 0 0 2 1 3 1 1 2 3 2 7 3 3 6
28-27 0 6 0 9 8 0 16 16 4 3 3 2 4 2 2 0 3 3 3 2 0 2 0
29-28 1 6 0 0 0 4 0 0 2 2 3 8 3 4 1 1 2 1 0 3 1 4 1
30-29 4 0 5 8 9 4 1 5 3 2 5 4 3 2 5 2 3 1 0 3 1 2 2
31-30 9 2 12 0 32 16 4 4 4 4 4 2 5 4 0 1 1 1 2 3 5 2 1
(Apr-19 onwards)

32-31 6 12 10 29 7 24 1 1 6 2 4 0 6 5 11 2 1 1 2 0 5 3 1
33-32 6 1 13 11 4 2 5 4 6 9 0 3 0 2 0 6 7 2 1 5 0 1 5
34-33 1 5 1 6 3 2 0 39 0 0 8 5 1 0 1 4 4 5 11 0 3 1 2
35-34 8 14 17 1 4 2 39 10 0 3 6 7 8 2 9 3 5 10 5 9 5 1 1
36-35 10 4 0 0 3 2 2 2 44 18 2 11 6 18 1 4 1 0 0 0 2 1 1
37-36 6 0 0 2 1 0 2 3 3 14 14 6 9 6 2 0 0 1 3 0 2 3 2
38-37 0 0 3 0 4 6 5 1 1 3 4 1 4 2 5 8 0 0 3 1 0 4 1
39-38 0 4 8 4 0 1 11 2 1 3 17 1 7 2 1 1 5 0 2 2 1 1 2
40-39 15 1 4 13 3 13 1 4 2 11 2 5 0 4 5 3 0 4 9 5 1 1 3
41-40 1 10 9 16 0 0 2 2 1 1 3 4 3 2 4 0 7 2 0 7 6 0 3
42-41 12 27 5 2 8 3 0 1 1 2 6 5 0 1 0 1 0 0 1 0 1 0 0
43-42 22 6 14 7 31 7 6 8 6 3 9 7 4 1 7 4 6 1 0 1 1 3 1
44-43 2 25 14 5 1 12 0 0 5 4 2 12 2 7 19 1 4 0 15 12 0 9 1
45-44 13 2 4 5 1 8 3 3 0 10 0 0 5 14 14 22 8 15 3 4 1 0 4
46-45 13 10 12 4 1 6 21 13 3 0 3 0 20 11 4 3 7 1 0 1 3 0 0
47-46 15 7 7 4 19 6 8 14 6 3 0 3 0 9 9 17 12 0 1 0 9 1 3
48-47 10 4 6 6 3 20 17 18 6 7 3 0 2 0 0 3 2 3 3 0 5
49-48 50 7 10 1 8 8 10 11 6 7 2 2 3 3 0 8 1 0 1 1
50-49 13 24 29 0 7 8 2 1 4 4 21 8 11 9 3 2 1
51-50 29 13 6 6 26 9 3 3 1 6 0 22 2 9 3 4 0
52-51 28 28 18 41 1 4 0 4 2 1 8 2 6 15 4
53-52 4 21 3 5 2 18 5 5 10 17 20
Table 5 Difference between VC of successive generators in SCED merit stack

54-53 20 2 18 20 17 4 15 5 6 2 1
55-54 19 17

Page 45
 Figure 14 Month-wise box plot of Variable Charges of SCED generators

Average Cost of Generation (Apr 2019 onwards)

240 Pre SCED Post SCED

230

220
Paise/kWh

210

200

190

180

Figure 15 Trend of daily average cost of generation under SCED

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 46

 4 LOOK AHEAD SCED

In the present SCED implementation, optimization is carried out one block at a time
with the objective of minimizing the total cost of dispatch. With this objective,

schedules of cheaper stations are pushed towards DC and those of costlier stations are
pushed towards Technical Minimum. As a result, the available ramp constrained

reserves get reduced when compared with those in pre-SCED schedules. When the
total change in ISGS schedules is greater than the ramp constrained reserves, it may

result in infeasibility in the optimization, which is resolved by adding artificial variables.

This aspect was mentioned in the detailed feedback report submitted to Hon’ble CERC

in January 2020.

It follows that if the SCED engine has visibility more number of time blocks at any

instant, it would be able to anticipate the expected ramp in the future blocks and

maintain adequate ramping capability. This should serve to reduce the number of
possible infeasibilities encountered in the SCED algorithm. In order to demonstrate

this, a look-ahead SCED algorithm has been developed and its operation and results
have been simulated for a period of one week (18-24 Jan 2021). The look-ahead

application is a dispatch-ready software with all operational features, including RRAS

dispatch. RRAS requirement in future blocks punched by shift engineers is also taken

into account in the model.

The look-ahead window chosen is 6 blocks, and total 8 blocks input data is taken,

including 2 previous-dispatched blocks. As most SCED generators are providing ramp

rate of 1%/min and the distance between DC and technical minimum is 45% of

capacity, 6 blocks is expected to provide sufficient look-ahead capability to the model.

Taking a very long look-ahead window may not provide greater benefit as the

schedules further down the line are liable to change due to revisions by constituents
and stations and the uncertainty increases as one looks further into the future. The

diagram at Figure 16 illustrates the approach adopted for look-ahead SCED.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 47

Run for Execution
95 96 1 2 3 4 5 6 7 8 9 10
block time

1 23:25 95 96 1 2 3 4 5 6

2 23:40 96 1 2 3 4 5 6 7

3 23:55 1 2 3 4 5 6 7 8

4 00:10 2 3 4 5 6 7 8 9

5 00:25 3 4 5 6 7 8 9 10

Old
Optimization window
Dispatch
Figure 16 Rolling look-ahead optimization window

The key results in terms of various optimization parameters and their comparison

with the results from the operational SCED implementation have been presented and
discussed below:

4.1 Infeasibilities
Net infeasibilities in the look-ahead model for one week period were -141 MWh,

due to which additional surplus of ₹ 3.7 Lakh would have been generated in the SCED
pool. For comparison, the net infeasibilities in the actual SCED dispatch over the same

period were -830 MWh, which resulted in additional surplus of ₹ 30 Lakh in the pool.
Thus, a significant improvement is seen with respect to quantum of infeasibilities and

the associated cost of relaxing these infeasibilities. Figure 17 shows the comparison of
blockwise infeasibility quantum in operational single period SCED vs simulated look-
ahead SCED.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 48

 Net Infeasibilities in SCED - Single period vs Look-ahead
MW
Single Period Look-ahead
400

300

200

100

-100

-200

-300

-400
18-01-21 19-01-21 20-01-21 21-01-21 22-01-21 23-01-21 24-01-21

Figure 17 Net Infeasibilities in SCED – Single period vs. Look-ahead

4.2 Cost Reduction

It is observed that the blockwise variable cost reduction in both the models is quite

close. At the same time, the look-ahead model is better able to avoid infeasibilities and
consequently, the savings spikes due to relaxation cost are also avoided. If the

relaxation cost component is excluded, the cost reduction in look-ahead SCED over
the week was ₹ 526 Lakh as compared to ₹ 512 Lakhs in single period SCED. The

comparison of blockwise cost reduction in SCED in both the models for sample days
is shown at Figure 18.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 49

 Cost Reduction in SCED - Single period vs Look-ahead Cost Reduction in SCED - Single period vs Look-ahead
10 Single period Look-ahead 4 Single period Look-ahead

5
0

-2
0
₹ Lakh/hour

₹ Lakh/hour
-4

-5
-6

-8
-10

-10

-15 -12
00:00
01:00
02:00
03:00
04:00
05:00
06:00
07:00
08:00
09:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00

00:00
01:00
02:00
03:00
04:00
05:00
06:00
07:00
08:00
09:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00
Figure 18 Cost reduction on sample days – Single period vs Look-ahead
RRAS Dispatch - Single period vs Lookahead RRAS Dispatch - Single period vs Lookahead
RRAS Punched RRAS-Single RRAS-Lookahead RRAS Punched RRAS-Single RRAS-Lookahead
1500 1500

1000 1000

500 500

0 0

-500 -500

-1000 -1000

-1500 -1500

-2000 -2000
00:00
01:00
02:00
03:00
04:00
05:00
06:00
07:00
08:00
09:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00

00:00
01:00
02:00
03:00
04:00
05:00
06:00
07:00
08:00
09:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00
Figure 19 RRAS dispatch on sample days – Single period vs Look-ahead

4.3 RRAS Dispatch

In terms of RRAS dispatch quantum vis-à-vis desired RRAS, the performance of

look-ahead model is better than single period model. This is because the look-ahead
model is able to incorporate change in RRAS requirement in the future and accordingly

adjust the dispatch in the current block in order to be able to meet this requirement
in future blocks. The average RRAS performance index (RRAS dispatch/RRAS desired)

in the look-ahead model was 92.4% which is higher than 89.4% in the single period
model. In most cases, lookahead model is able to provide the desired RRAS quantum,

except in some blocks with insufficient spinning reserves. Figure 19 shows plots of
desired RRAS and RRAS dispatched with single period and lookahead models for
sample days.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 50

 Ramping Up Reserves Ramping Down Reserves
MW MW
Single period Look-ahead Single period Look-ahead

5000 8000

4500
7000

4000
6000
3500
5000
3000

2500 4000

2000
3000

1500
2000
1000

1000
500

0 0

Figure 20 Ramp constrained reserves on sample day – Single period vs Look-

ahead

MW Ramping Up Reserve - Duration MW Ramping Down Reserve - Duration

6000 Single period Look-ahead 8000 Single period Look-ahead
7000
5000
6000
4000
5000

3000 4000

3000
2000
2000
1000
1000

0 0
0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%

Figure 21 Duration curve of ramp constrained reserves over study period

4.4 Availability of Ramp-constrained reserves

As mentioned at the beginning of this chapter, under single-period SCED, ramp
constrained reserves get reduced when compared to beneficiaries’ schedules.

Theoretically, look-ahead SCED should enhance ramp-constrained reserves in the

blocks where it is needed to accommodate future ISGS demand ramp and thus reduce

occurrence of probable infeasibilities. Figure 20 Figure 19 shows the ramp-constrained

reserves over 96 blocks with both the models on a sample day. Figure 21 shows

duration curve of ramp constrained reserves with both the models over one week. It is
apparent that ramping up reserves are slightly higher with the look-ahead model in

order to meet the ramp in ISGS schedules. Ramping down reserves are sufficient during
the week to meet the ramp down in total ISGS schedules, hence there is not much
need to increase them. Further, it can be seen that for approximately 10% of time

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 51

during the week, ramping up reserves are zero even with the look-ahead model. This
is owing to exhausted spinning reserves in ISGS. Introduction of unit commitment with

SCUC would help in maintaining sufficient spinning reserves.

4.5 Ease of generator operations

Ease of operations of generators is assessed in terms of number of ramp up or
down instructions and cumulative MW change in schedules. On All India basis for SCED

generators in the week under study, the number of instructions in the look-ahead
output schedule reduced by 32% from pre-SCED schedule, while in operational single

period SCED, this reduction was 35%. Cumulative MW change on all India basis in look-
ahead output schedule reduced by 9.4% from pre-SCED schedules, compared to 6.8%

reduction in single period SCED. Figure 22 shows the station-wise pre-SCED

cumulative MW change in schedule along with percentage change in look-ahead and

single period SCED.

% Change in Cumulative MW Change w.r.t. Pre-SCED Schedule

600 Pre SCED Cum. Change (on sec. axis) % Change - Lookahead % Change - Single period 45000

500 40000

35000

Pre-SCED Cumulative change (MW)

400
% change from Pre-SCED

30000
300
25000
200
20000
100
15000

0
10000

-100 5000

-200 0
Korba STPS Stage 1 and 2

Ramagundam TPS Stage 3

Ramagundam TPS Stage 1&2

NTPL
MPL
Sasan

Barh

Kudgi-1
Vellur NTECL

Dadri Stage 1

Dadri Stage 2
Farakka Stage 1 and 2
Darlipali

Rihand 3

Rihand 2
Rihand 1
Singrauli TPS

Vindhyachal STPS Stage 4

Vindhyachal STPS Stage 3
Vindhyachal STPS Stage 2
Vindhyachal STPS Stage 5
Vindhyachal STPS Stage 1

Kahalgaon Stage 1
Nabinagar

Unchahar TPS 1
Talcher
LARA

NSPCL
Kahalgaon Stage 2
NPGC

Tanda-2
Sipat STPS Stage 1
Sipat STPS Stage 2

Korba-Stage-3

FARAKKA-STAGE-3

Mauda

Bongaigaon
Simhadri -NTPC Stage -1

Unchahar TPS 3
Unchahar TPS 2

NTPC SOLAPUR
Gadarwara-1

UNCHAHAR 4

MTPS Stg 2

Mauda 2
Talcher Stage 2-NTPC

NTPC-KHARGONE
Simhadri NTPC Stage 2

Indira Gandhi TPS (Jhajjar)

Figure 22 Ease of generator operations with SCED – Single period vs Look-

ahead

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 52

4.6 Schedule variability within look-ahead window
It has been observed during the operation of the SCED pilot that, in general, the

shape of the total ISGS schedule curve over 96 blocks at the end of the day is
significantly different from the day-ahead schedule. Usually, the day-ahead schedule

is relatively flat before the start of the day and the variability gets incorporated due to
revisions by constituents closer to real time based on individual state portfolio

balancing. Figure 23 shows the comparison of total schedule of SCED and RRAS
generators at the start and end of day for a sample day.

All India ISGS Schedule - Day-ahead vs final revision

51000
Day ahead Schedule Final Schedule
50000

49000

48000

47000

46000

45000

44000

43000
1-Mar-21
42000

Figure 23 Change in All India ISGS schedule over day

Energy Analytics Lab, IIT Kanpur recommended a rolling block implementation

framework in their paper10. The paper recommended keeping a window of 4-6 time

blocks and concluded that a longer window was not resulting in more savings.
However, the paper took only the final schedule at the end of the day as input and did

not consider the schedule variability mentioned above.

10 A. Singh, T. B. Kumar, G. Yadav and R. Karna, "Security Constrained Economic Despatch – India: A Rolling
Block Implementation Framework," 2019 8th International Conference on Power Systems (ICPS), Jaipur, India,
2019, pp. 1-6, doi: 10.1109/ICPS48983.2019.9067641.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 53

 SCED implementation working in real-time would be able to incorporate the future
schedule visible at that point in time, which is liable to change in subsequent blocks.

For example, a ramp event visible 6 time blocks ahead may not materialize when this
block’s schedule is finalized after RTM clearing. Hence, a choice of look-ahead window

longer than 6 blocks has a lot of uncertainty and variability associated with it.

Several visualizations have been attempted to represent this uncertainty aspect in

total ISGS schedule (which serves as proxy for demand for SCED). Some of these
visualizations have been presented for a sample day.

Net ISGS schedule in look-ahead window

52000

50000

48000

46000
Total Schedule (MW)

44000

42000

40000

38000

36000

34000

32000
1 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61 65 69 73 77 81 85 89 93 97 101
Block number

Figure 24 Net ISGS schedule within look-ahead run window

Figure 24 shows the block-wise net ISGS schedule for each run of look-ahead SCED.
Over the course of a day, there are 96 SCED runs, and each of these runs has a visible

window of 6 blocks, which get shifted by one block in every run. Each line segment in
the plot represents one SCED run window (6 blocks look-ahead). Joining the first point

in each of these segments would give the final ISGS schedule implemented at the end
of the day. The variability across these run windows is due to impact of revisions by

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 54

constituents and RTM. If the final revision is taken as the input for look-ahead SCED,
all of the line segments would overlap with each other and with the final ISGS schedule

curve. It needs to be noted that RTM schedules do not factor ramping. If RTM has to
consider this aspect, then PX must get ISGS schedule data from RLDCs/NLDC through

API and change their algorithm. That still leaves a gap if a generator bids in both
exchanges, which is quite unlikely.

Variability in block-wise total schedule from T-8 to T-3 (final dispatch)

Difference T-3 (final) T-4 T-5 T-6 T-7 T-8
52000 3500

50000 3000

Difference between total schedule at T-8 and T-3 (MW)

48000 2500

46000 2000
Total Schedule (MW)

44000 1500

42000 1000

40000 500

38000 0

36000 -500

34000 -1000

32000 -1500
1 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61 65 69 73 77 81 85 89 93
Block number

Figure 25 Variability in block total schedule visibility over time

Figure 25 attempts to show this variability through another approach. As SCED is

being dispatched on T-3 basis, any particular block enters the look-ahead window in

the SCED run in T-8 block and finally gets dispatched in the SCED run in T-3 block. The

visualization shows the total ISGS schedule for a particular block between these points
in time and bars denote the change in the total schedule from T-8 to T-3.

Figure 26 depicts the variation of blockwise lookahead schedule over different runs.
The horizontal axis represents the sequential SCED run number for the day. Each line
segment represents the total schedule for one block over subsequent runs. The first
point in each line segment is at run T-8 and the final point is at run T-3 (final dispatch).

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 55

If there were to be no schedule revision within the look-ahead window, each line
segment would be a horizontally flat line. The solid red line joining the segments

represents the final ISGS dispatch schedule at the end of the day.

Variation of blockwise total schedule over different runs

52000

50000

48000

46000
Total Schedule (MW)

44000

42000

40000

38000

36000

34000

32000
1 7 13 19 25 31 37 43 49 55 61 67 73 79 85 91 97
Run number (Block number of dispatch block)

Figure 26 Variation of block total ISGS schedule across runs

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 56

5 SECURITY CONSTRAINED UNIT COMMITMENT

5.1 Motivation
Presently unit commitment for ISGS units is indirectly done by the constituents

based on their requirement inline with the CERC approved Reserve Shutdown (RSD)
procedure11. If the total day ahead requisitions fall below technical minimum for 8

continuous hours, then the unit may get decommitted by the station. The main
drawback of this approach is:

1. Reserves are not factored in while deciding the unit commitment. This leads to
depletion of spinning reserves especially during peak hours.

2. Constituents with small share in the unit are unable to exert claim on the unit

commitment. When all units in a plant go under RSD the offbar requisition is
not honoured despite the constituents paying fixed charges.

3. The unit commitment decisions are at times sub optimal due to the block
diagonal structure of the ISGS share allocation matrix. This leads to cheaper

units in one region under reserve shutdown and costlier units in another region
remaining on bar.

Good utility practice dictates that unit commitment is precluded by a reliable

short-term demand and RE forecast. To address these issues, it is proposed to run

Security Constrained Unit Commitment (SCUC) centrally at NLDC after the activities at
RLDCs in-line with the RSD procedure. Similar to the SCED framework, SCUC will be a

thin layer of optimization over the existing decentralized RSD procedure. The unit
commitment decided by the existing RSD procedure will be provisional. This is similar

to the provisional ISGS schedules prepared by RLDC in the SCED framework. Then
SCUC will take an All India view and provide the final unit commitment keeping in view

11Hon’ble CERC approved detailed procedure for taking unit(s) under Reserve Shut Down and Mechanism for
Compensation for Degradation of Heat Rate, Aux Compensation and Secondary Fuel Consumption, due to Part
Load Operation and Multiple Start/Stop of Units (http://www.cercind.gov.in/2017/regulation/SOR132.pdf)

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 57

system security and economy. This step is similar to the final optimized ISGS schedules
prepared by NLDC.

Figure 27 shows the 2019-20 un requisitioned surplus (URS) duration curve of the
ISGS which indicates that reserves are available for 60-70% of the time. For the

remaining 20-30% of the time also, maintaining spinning reserves would improve the
reliability of the Indian power system. Reserves in 2020-2021 have not been analysed

as it is an abnormal year considering COVID-19 impact on the power system.

Figure 27 Regulation quantum duration available during 2019-20 (MW)

Further, the duration curve of the difference of sell bids and buy bids in the Power
Exchange (IEX) is shown in Figure 28. From this, it is observed that around 10% of the

time, buy bids are more than sell bids implying that enough generation is not available
in the day-ahead energy market also during this period. Such periods are generally

observed to be falling during the peak hours. In such cases, availability of regulating
reserves in the market would also be a challenge.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 58

 Figure 28 Surplus of sell bids over buy bids in Day-Ahead Market in Power
Exchange

5.2 SCUC Formulation

SCUC being an optimization problem has an objective function and a set of

constraints to be honoured. The objective is to minimize the ISGS operation cost for
the time window under consideration. The time window should be at least 3 days to

enable starting of units with a cold startup time of 40 hours. A 3-day window can be
chosen in line with the draft IEGC 2020. It is however ideal to use a 7-day window due

to its periodic nature. The constraints are to honour system security and equipment
limitations. A detailed literature survey was also conducted and it is summarized at

Annexure-5.

5.2.1 Objective function

The objective of SCUC is to minimize the 3 day ahead ISGS operation cost. The
operation costs comprise of energy cost and startup cost. Heat rate degradation has

been neglected to keep the formulation in the linear domain.

5.2.2 Constraints

1. Meeting the 3 day ahead forecasted demand:

The 3 day ahead ISGS demand will be forecasted by NLDC based on the following
inputs

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 59

 (a) Demand forecast

(b) RE forecast

(c) Hydro forecast

(d) Day Ahead Market price

(e) Day ahead ISGS availability

This would serve as the demand proxy for SCUC.

2. Maintaining the required spinning reserve:

The spinning reserve required to be maintained will be entered by the NLDC

operator. Reserves take care of the real time uncertainties like unit tripping, unit
derating and errors in demand forecast, weather events. This would be

configurable with default of 3 GW and a provision for region wise spinning

reserves.

3. Honoring transmission constraints:

SCUC will ensure that the unit commitment honors the ATC constraints.

4. Operator intervention like Must RUN/ Must OFF units:

This provision allows the operator to override the unit commitment decision of

SCUC especially to ride through constraints

5. Capacity and Ramp constraints:

The DC, technical minimum and ramp rates declared by the plants will be
honoured while deciding the unit commitment.

6. Minimum Up Time and Minimum Down Time:

When a unit is committed, it has to remain ON for certain minimum time called
Minimum Up Time. Similarly, when a unit is decommitted it has to remain OFF

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 60

 for certain minimum time called Minimum Down Time before it can be brought
back on bar.

7. Crew constraint:

Typically, the number of crew in a power plant will be just sufficient to start one
unit at a time. This constraint ensures a minimum time gap between startup of

two units in the plant. Since all the units in a stage will have the same variable
cost, SCUC would commit them at the same time if crew constraint is not

modeled.

The inputs required by SCUC are listed in Table 6.

Table 6 Inputs required for SCUC

S. no. Input Entities
1 Blockwise All India and Regional ISGS demand forecast for NLDC
next 3 days
2 Blockwise spinning reserve requirement for next 3 days NLDC
3 Blockwise ATC margins for next 3 days NLDC
4 Operator decided unit commitment NLDC
5 Blockwise unit availability, unit DC, unit Technical Minimum ISGS
and unit ramp for next 3 days
6 Unit Minimum Up Time and Minimum Down Time ISGS
7 Crew constraint time ISGS
8 Energy cost and startup costs (Hot/Warm/Cold) ISGS
9 Startup time (Hot/Warm/Cold) and unit boiler status ISGS

5.3 Dovetailing SCUC to the existing framework

The proposed SCUC timelines are shown in Table 7

Table 7 Proposed scheduling activities and timelines for SCUC

S. no. Time Activity (Day Ahead) Entities

1 06:00 Unit DC summed up to obtain Plantwise DC ISGS

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 61

 S. no. Time Activity (Day Ahead) Entities
2 08:00 Plantwise Entitlement published RLDC
3 09:30 Consent to ISGS for selling in DAM Constituents
4 15:00 R0 requisition punched Constituents
5 18:00 R0 schedules prepared RLDC

6 20:00 RSD procedure provides a provisional unit RLDC

commitment
7 21:0012 SCUC provides the binding unit commitment NLDC

5.4 Changes to Web Based Scheduling Software (WBES)

Some changes would be needed in WBES for implementation of SCUC. The draft

new practices, needed to be adopted by power stations, have been listed below.

5.4.1 Unit DC

In order to run SCUC unit wise DC is required. Presently ISGS provide only
station DC. It is difficult to estimate the unit DC from station DC when different unit

sizes are present in the same station. Example: A station (3×500 MW) provides a
derated DC of 1200 MW. There are 2 possibilities for the unit DC

(a) 3 units have a derated DC of 400 MW each

(b) 2 units have a DC of 500 MW and the third unit is in the middle of

synchronization with a DC of 200 MW.

Hence stations are required provide unit wise DC that will be summed up to arrive
at plant DC. Scheduling will continue to be station wise. Constituents would continue
to requisition plant wise and not unit wise. This will be an internal parameter in WBES

only for the purpose of SCUC.

Stations would need to declare DC three days in advance for SCUC. SCUC

software would get details of power plants viz., DC, DC-onbar, DC for schedule, Ramp

12 A rolling 3-day SCUC would ensure sufficient advance notice for committing units and therefore day-ahead
run at 21:00 hrs would not be too late.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 62

rate from WBES.

5.4.2 Pmax in WBES and SCED

The Pmax in WBES will depend on the unit commitment of the RSD procedure
and in SCED will depend on SCUC.

WBES Pmax = ∑𝑢𝑛𝑖𝑡𝑠 𝑅𝑆𝐷 𝑆𝑡𝑎𝑡𝑢𝑠 𝑋 𝑈𝑛𝑖𝑡 𝐷𝐶

SCED Pmax = ∑𝑢𝑛𝑖𝑡𝑠 𝑆𝐶𝑈𝐶 𝑆𝑡𝑎𝑡𝑢𝑠 𝑋 𝑈𝑛𝑖𝑡 𝐷𝐶
The RLDC schedule will be clamped at WBES Pmax when requisitions

(onbar+offbar) go beyond WBES Pmax.

5.4.3 Pmin in WBES and SCED

The Pmin in WBES will depend on the unit commitment of the RSD procedure
and in SCED will depend on SCUC. The RLDC schedule will be jacked up to WBES Pmin

in case requisitions fall below WBES Pmin. The RLDC schedule will be clamped at WBES
Pmax when requisitions (onbar+offbar) go beyond WBES Pmax.

5.4.4 Requisitions by Constituents

States have to provide requisition based on the DC three day in advance. States

may be allowed to change the requisition any time as per their requirements. SCUC
will fetch requisition details of beneficiary of the power plants from WBES.

5.4.5 Unit De-Commitment through SCUC

Withdraw the unit as per ramp rate, in line with SCUC advisory to de-commit

the units. Honour the requisition of the beneficiary having share allocation in the
power station, which got de-committed due to SCUC.

i. An option would be made available in power station to participate under

SCUC. If the plant choses to participate in SCUC, option for plant “To fill Off

bar SCUC DC” shall be made available.

ii. Calculate On-Bar DC = DC for schedule - Off bar SCUC DC

iii. NLDC shall advice power station to withdraw the Unit on account of SCUC.
iv. For withdrawal of Unit, code may be exchanged among NLDC/RLDC/Power

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 63

 station. This information shall be later used for SCUC settlement.
v. Beneficiaries shall be allowed to give requisition based on On-Bar DC.

vi. Scheduling software shall prepare the schedule based on the requisition Up
to limit of On-Bar DC as per current practice.

vii. SCED shall provide the SCED down to power station getting schedule above
the On-Bar DC.

5.4.6 Unit Commitment through SCUC

Plant shall revise the On Bar DC as per the ramp rate, in line with the SCUC

advisory (to commit the units). SCED will check if schedule of Plant is equal to technical

minimum considering the new on bar DC.

a. If yes, support from SCED would not be there.

b. If No, SCED will send positive SCED to provide technical minimum support.
For synchronizing of unit under SCUC, code may be exchanged among

NLDC/RLDC/Power station. This information shall be later used for SCUC settlement.
Beneficiary shall be allowed to give full requisition on On-Bar DC. Any requisition

below the technical minimum shall be supported by SCED duly considering any
generation that might have got decommitted.

5.5 SCUC Case Study

ISGS SCUC was run for an eight-month period from June 1, 2020 to Jan 31, 2021.

Each run was for 3-day window on a rolling basis. The run for the period, say (Sep 15
to Sep 17) would give binding unit commitment decisions for Sep 15 and provisional

unit commitment decisions for Sep 16 and 17. The starting point for the next run (Sep
16 to Sep 18) would be obtained from the (Sep 15 to Sep 17) run.

5.5.1 Inputs and Assumptions

The inputs and assumptions are as follows:

 Final SCED optimized ISGS schedules has been used to arrive at the proxy
demand.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 64

  Total All India Spinning Reserve requirement of 3 GW in every block. This
constraint was relaxed if it proved infeasible.

 Actual ATC margins used for constraining regions

 Block wise Unit Availability derived from the plant Onbar Normative DC

and Offbar DC.

 Actual Variable cost for all units.

 Actual Technical Minimum percentage for all units.

 Assumed ramp of 1% /min for all units.

 Startup costs has been assumed to comprise only of the oil cost. The oil

consumption for startups has been taken as per IEGC. The cost of oil
assumed as 40 Rs/l.

 Minimum Up Time and Minimum Down Time assumed a 8 hours in line

with RSD procedure.

 Crew constraint time assumed as 5 hours.

All assumed inputs and configurable.

5.5.2 Classification of Units in SCUC

After SCUC runs, the available units can be divided into 5 groups:

1. Initially ON and committed by SCUC (Quadrant 1 (blue)).

2. Initially ON and decommitted by SCUC (Quadrant 2 (green)). It would be

suboptimal to continue running these units and hence get decommitted by

SCUC.

3. Initially under RSD and decommitted by SCUC (Quadrant 3 (violet)).

4. Initially under RSD and revived by SCUC (Quadrant 4 (brown)). It would be

suboptimal to continue keeping these units under RSD and hence get
committed by SCUC.

5. Two shifting unit (Red quadrant). A two shifting unit has at least one startup
and one shutdown in the same day.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 65

 The All India and region wise classification of the 131 available units for SCUC run
(01-Jun to 03-Jun) is shown in Figure 29.

Initially on and Initially on and

decommitted by SCUC committed by SCUC

Initially under RSD and Initially under RSD and

decommitted by SCUC revived by SCUC

Figure 29 Classification of units post SCUC

The cheapest and costliest unit within each group is shown in Table 8.

Table 8 Cheapest and Costliest unit within each group

S no. Group Cheapest (VC) Costliest (VC)

1 Initially ON and committed by SCUC Darlipalli U1 (1.08) Barh U4 (3.03)
2 Initially ON and decommitted by SCUC Barh U5 (3.03) Dadri U1 (3.85)
Initially under RSD and decommitted by
3 Mauda U1 (2.86) Dadri U2 (3.85)
SCUC
4 Initially under RSD and revived by SCUC BRBCL U3 (2.39) Khargone U1 (3.10)
5 Two shifting units NIL NIL

All units affected by SCUC (Quadrants 2 and 4) are listed in Table 9 to

reemphasize the savings that can be accrued due to regional diversity.

Table 9 Units impacted by SCUC (Quadrants 2 and 4)

Quadrant 2 (green) Quadrant 4 (brown)

S no. Unit MW VC Unit MW VC
1 Barh U5 660 3.03 BRBCL U3 250 2.39
2 Unchahar U6 500 3.43 NSPCL U2 250 2.59

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 66

 Quadrant 2 (green) Quadrant 4 (brown)
S no. Unit MW VC Unit MW VC
3 BGTPP U1 250 3.46 Mauda U2 500 2.86
4 BGTPP U2 250 3.46 Mauda U3 660 2.87
5 BGTPP U3 250 3.46 Mauda U4 660 2.87
6 Vallur U1 500 3.56 Gadarwara U1 800 2.93
7 Vallur U2 500 3.56 Simhadri U1 500 3.02
8 Vallur U3 500 3.56 Khargone U1 660 3.10
9 Dadri U5 490 3.63
10 Unchahar U1 210 3.64
11 Unchahar U5 210 3.64
12 Unchahar U3 210 3.67
13 Unchahar U4 210 3.67
14 Dadri U1 210 3.85
Total 4950 Total 4280

The revived capacity is only 4280 MW against the decommitted capacity of 4950

MW, as during 01-Jun to 03-Jun more reserves were present than required as shown
in Figure 30. SCUC maintains the required amount (not excess not less) optimally.

Figure 30 Spinning Reserves comparison from 01 Jun to 03 Jun

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 67

 5.5.3 Impact on Reserves

The breakup of the installed capacity post SCUC is shown in Figure 31. The

monsoon to winter transition is clearly visible with the Demand increasing and cold
reserves decreasing.

Figure 31 Seasonal Transition of Reserves

The comparison of the ramp constrained reserves is shown in Figure 32. The
comparison of the spinning reserves is shown in Figure 33.

Summary of results over the 8 month period is shown in Table 10.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 68

 Figure 32 Ramp-constrained Reserves comparison

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 69

 Figure 33 Spinning reserves comparison

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 70

 Table 10 SCUC Cost Summary (Jun 2020 – Jan 2021)

S no. Scenario Energy Cost (₹ Cr) Startup Cost (₹ Cr) Savings (Cr)
1 Pre SCED 44,731 + SCED - -
2 Post SCED 44,731 - SCED
3 SCUC with 3 GW 44,227 113 391+SCED
reserve
4 SCUC without reserve 44,014 81 636+SCED

The heat map of the SCUC System Marginal Price (SMP) is shown in Figure 34.

During monsoon the SMP is highest during the evening peak and lowest during solar
period, whereas in winter it is highest during the morning peak and lowest during off-

peak period.

The plot of SMP and the VC of the cheapest unit providing cold reserve is shown

in Figure 35. Most of the time SMP is below the cheapest cold reserve validating the
optimality of SCUC results.

Figure 34 SMP Heat Map

Figure 35 SMP vs Cheapest Cold Reserve

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 71

 A plot of daily additional savings over and above SCED due to SCUC and startup
costs is shown in Figure 36.

Figure 36 Daily additional savings due to SCUC and startup costs

5.5.4 Change in schedule due to SCUC+SCED

At all India level, the change is net schedule is zero, indicating there was no deficit/
excess energy post SCUC (Figure 37). Net perturbation due to SCUC in different

regions over the 8 month period is: ER: -132 MU (Figure 38), NER: -1406 MU (Figure
39), NR:-4133 MU (Figure 40), SR: -1122 MU (Figure 41), WR: +6794 MU (Figure 42).

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 72

 Figure 37 All India SCUC Schedule vs Actual Schedule

Figure 38 ER SCUC Schedule vs Actual Schedule

Figure 39 NER SCUC Schedule vs Actual Schedule

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 73

 Figure 40 NR SCUC Schedule vs Actual Schedule

Figure 41 SR SCUC Schedule vs Actual Schedule

Figure 42 WR SCUC Schedule vs Actual Schedule

Plots for stations with high perturbation under SCUC are shown below. These are

IGSTPS Jhajjar: -1333 MU (Figure 43), Dadri-II: -1106 MU (Figure 44), Solapur: +1559
MU (Figure 45), Mauda-II: +1496 MU (Figure 46).

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 74

 Figure 43 IGSTPS Jhajjar SCUC Schedule vs Actual Schedule

Figure 44 Dadri-II SCUC Schedule vs Actual Schedule

Figure 45 Solapur SCUC Schedule vs Actual Schedule

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 75

 Figure 46 Mauda-II SCUC Schedule vs Actual Schedule

5.5.5 Unit Status

The status of a unit can be

Outage: The unit is unavailable

RSD: The unit is available but decommitted

ON: The unit is available and on bar

The actual and SCUC unit status for June 2020 to August 2020 is shown in Figure 47.
Being wind season this period saw many units in RSD. Due to regional diversity, there

are some relatively cheaper units under RSD in Actual status. But SCUC optimizes and
ensures only costly units in RSD.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 76

 Figure 47 Unit status comparison for June 2020 to August 2020

The SCUC unit status for December 2020 to January 2021 is shown in Figure 48. No

units are kept continuously in RSD and there is less room for optimization. In actual
status, 4 units of Dadri-I are sub optimally under RSD. SCUC revives all units in Dadri-

I to provide reserves.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 77

 Figure 48 Unit status comparison for December 2020 to January 2021

5.5.6 Impact of SCUC on startups

Startups are classified as hot, warm, cold and non SCUC. A non-SCUC startup occurs
when a unit is revived from forced outage/ planned maintenance. The startup costs for

non-SCUC startups are to be borne by the plant and are not reflected in savings
calculation. Figure 49 shows the scatter plot between unit VC and number of startups

during the 8-month period.

There are 0 hot starts, 215 warm starts, 212 cold start and 573 non-SCUC starts in the

8-month period. Most of the warm and cold starts occur for units whose VC is in range
2.8 Rs/kWh to 3.5 Rs/kWh.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 78

 Figure 49 Startups Scatter plot

The heatmap showing diurnal startup trend is shown Figure 50. During monsoon most
the startups occur during the evening peak in contrast to winter where they occur in
morning peak.

Figure 50 Diurnal startup pattern

SCUC software will fetch details of Cold/Warm/Hot start up time from RRAS AS3
format. In order to keep record of start /stop due to SCUC, details shall be logged at
NLDC/RLDCs to avoid frequent start stop of Unit on account of SCUC.

5.6 Optimization Formulation of SCUC

SCUC being an optimization problem has an objective function and a set of constraints

to be honoured. The objective is to minimize the operation cost for the time window
under consideration. The constraints are to honour system security and equipment
limitations.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 79

For the purpose of simulation and implementation of SCUC, apart from the core
algorithm, a web-based application has been developed internally. The salient features

of this application are detailed at Annexure-6.

System Security Constraints

 Meeting Demand
 Ensuring spinning reserves

 ATC constraints
 Operator intervention like must ON/ must OFF

Equipment constraints

 Unit availability
 Capacity and Ramp constraints of units

 Minimum Up/Down times for the units

 Initial Up/Down times for the units

 Crew constraint for plants

5.6.1 Notation

Sets

 i : Generator index

 p : Plant index
 t : Time block

 r : Region index

Variables

 ui,t : Status variable (1 if unit is ON at time t , 0 otherwise)

 Pi,t : Optimal dispatch setpoint unit

 si,t : Start up variable (1 if unit is turned ON at time t, 0 otherwise)

 zi,t : Shut down variable (1 if unit is turned OFF at time t, 0 otherwise)

Parameters

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 80

  ai,t : Unit availability (1 if unit is available at time t, 0 otherwise)
 𝐶𝑖̅ : Pmax (Normative capacity)

 Ci : Pmin (Technical Minimum)

 RUi : Ramp Up limit

 RDi : Ramp Down limit

 UTi : Minimum Up Time

 DTi : Minimum Down Time

 IUTi : Initial Up Time

 IDTi : Initial Down Time

 Tp : Crew constraint time for plant p

 ci : Variable charge

 ST i : Cold Start Up cost

 Dt : Demand to be met

 𝐿̅𝑡 : Spinning UP Reserve requirement

 IMr,t : ATC Import margin at the time of SCUC run

 EMr,t : ATC Export margin at the time of SCUC run

5.6.2 Objective function

The objective of SCUC is to minimize the 3 day ahead operation cost. The operation
costs comprise of energy cost and startup cost. Heat rate degradation is has not been

considered to keep the formulation in the linear domain. The startup cost is assumed
to be independent of the type of start (Hot/Warm/Cold). Cold startup cost was chosen

to penalize each startup to reduce unit toggling. A 3-day window was chosen in line
with the draft IEGC 2020. It is ideal to use a 7-day window due to its periodic nature.

The operation cost is mathematically written as:

Operation Cost = ∑ ∑( 𝑐𝑖 𝑃𝑖,𝑡 /4 + 𝑆𝑇𝑖 𝑠𝑖,𝑡 ) + ∑ 𝑀 × 𝐴𝑟𝑡𝑡 /4

𝑡 𝑖 𝑡

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 81

The first term is the energy cost associated with unit i for time t. The second term is
the startup cost associated with unit i for time t. The above terms are summed for all

units and over all time.

5.6.3 Startup and Shutdown Variables

The formulation of minimum Up/Down time constraint requires to define exclusive

variables which capture the unit’s transition from OFF to ON and vice versa. Variables

which capture the startup and shut down of the units are denoted as si,t and zi,t
respectively.

si,t : Start up variable (1 if unit is turned ON at time t, 0 otherwise)

zi,t : Shut down variable (1 if unit is turned OFF at time t, 0 otherwise)

Figure 51 graphically shows the relation between the status variable ui,t and the startup

and shutdown variables. We can also deduce a mathematical equation relating the
status variable and startup/shutdown variables.

ui,t − ui,t−1 = si,t − zi,t (6.1)

Eqn. (6.1) allows a possibility of si,t = zi,t = 1, when the unit remains ON or remains OFF.

To prevent this the constraint Eqn (6.2) is additionally imposed.

si,t + zi,t ≤ 1 (6.2)

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 82

 Figure 51 Startup and Shutdown variables

5.6.4 Constraints

Meeting the Demand

Since SCUC will be run only for ISGS thermal generators, the 3 day ahead requirement

from these plants would serve as proxy demand. Eqn. (6.3) will ensure that generation
is at least the proxy demand for all time blocks

∑𝑖 𝑃𝑖,𝑡 ≥ 𝐷𝑡 (6.3)

Ensuring spinning reserves

The spinning reserve present in each unit is the difference between its on bar
normative capacity and its dispatch setpoint. Eqn. (6.4) ensures total spinning reserve

must be more than the operator entered reserve requirement.

∑ (𝑢𝑖,𝑡 𝐶𝑖 − 𝑃𝑖,𝑡 ) ≥ 𝐿𝑡 − 𝐴𝑟𝑡𝑡 (6.4)

𝑖

The first term inside the summation is the on bar normative capacity and the second
term is the optimal dispatch setpoint.

ATC Constraints

SCUC should ensure ATC margins are honoured while deciding the unit commitment.

Eqn. (6.5) ensures the power backed down in each region should be less than the ATC
Import margin. Eqn. (6.6) ensures the power increased in each region should be less

than the ATC Export margin. Change in ISTS losses is ignored for simplicity.

𝐷𝑟,𝑡 − ∑𝑖∈𝑟 𝑃𝑖,𝑡 ≤ IM𝑟,𝑡 (6.5)

∑𝑖∈𝑟 𝑃𝑖,𝑡 − 𝐷𝑟,𝑡 ≤ EM𝑟,𝑡 (6.6)

Operator Intervention

This constraint accommodates the operator’s discretion in SCUC. This can be used by
the operator to commit a unit which otherwise would be decommitted by SCUC and
vice versa. For this we have a must RUN flag and must OFF flag.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 83

 ui,t = ai,t if unit i must RUN flag = 1 (6.7)

ui,t = 0 if unit i must OFF flag = 1 (6.8)

Eqn. (6.7) and Eqn. (6.8) explicitly force SCUC to use the operator specified unit
commitment.

Unit Availability

Units may be unavailable due to forced outages or planned maintenance. Such units

do not participate in SCUC. Eqn. (6.9) ensures the unit is OFF when availability entered
is zero.

ui,t ≤ ai,t (6.9)

Capacity constraints

When a unit is committed SCUC should ensure its optimal dispatch setpoint to be

between its normative capacity and technical minimum. This is enforced by Eqn. (6.10)
and (6.11).

Pi,t ≤ ui,t𝐶𝑖̅ (6.10)

Pi,t ≥ ui,t Ci (6.11)

Ramp Constraints

The difference between unit’s optimal dispatch setpoint for any two consecutive blocks

should be within its ramp capability. The ramp constraint Eqn. (6.12) also ensures a
smooth startup by first operating at techmin for the first block after startup. Similarly

with Eqn. (6.13) smooth shutdown is ensured by bringing the unit to techmin before
shutting it down.

Pi,t − Pi,t−1 ≤ ui,t−1 RUi + si,t Ci (6.12)

Pi,t−1 − Pi,t ≤ ui,t RDi + zi,t Ci (6.13)

Minimum Up time

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 84

When a unit is committed it has to remain ON for certain minimum amount of time
called minimum up time. Eqn. (6.14) enforces this constraint.
𝑡
∑𝑘=𝑡−UT +1 𝑠𝑖,𝑘 ≤ 𝑢𝑖,𝑡 (6.14)
𝑖

The LHS in Eqn. (6.14) is the number of startups in the preceding UTi time window. This

constraint implies the following:

 If unit is ON at t, it can turn ON atmost once in the preceding UT window.

 If unit is OFF at t, it cannot turn ON in the preceding UT window. Consider case

when ui,t = 1.

Consider case when ui,t = 1.

(a) (b)

Figure 52 Illustration of Minimum up time (a)Valid (b)Invalid

In plot shown in Figure 52 (a), the unit turns ON in the window and remains ON till the

end of the window. Hence no violation of minimum Up time. But in the plot shown in
Figure 52 (b) the unit turns ON twice in the window violating the minimum Up time.

Consider case when ui,t = 0.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 85

 Figure 53 Illustration of Minimum up time (a)Valid (b)Invalid

In plot shown in Figure 53 (a), the unit does not turn ON in the window and is OFF at
the end of the window. Hence no violation of minimum Up time. In the plot shown in

Figure 53 (b), the unit turns ON inside in the window but is OFF at the end of the
window. Hence violation of minimum Up time is there.

Minimum Down time

When a unit is decommitted it has to remain OFF for certain minimum amount of time

called minimum down time. Eqn. (6.15) enforces this constraint.

𝑡
∑𝑘=𝑡−DT +1 𝑠𝑖,𝑘 ≤ 1 − 𝑢𝑖,(𝑡−𝐷𝑇𝑖 ) (6.15)
𝑖

The LHS in Eqn. (6.15) is the number of startups in the preceding DTi time window. This
constraint implies the following

 If unit is ON at t, it cannot turn ON in the succeeding DT window.

 If unit is OFF at t, it cannot turn ON once in the succeeding DT window.

Consider case when ui,(t−DTi) = 1.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 86

 Figure 54 Illustration of Minimum down time (a)Valid (b)Invalid

In plot shown in Figure 54(a), the unit does not turn ON in the window and is fully OFF
in the window. Hence no violation of minimum Down time. In the plot shown in Figure
54(b), the unit turns ON inside in the window, its OFF time is less than DTi. Hence

violation of minimum Down time.

Consider case when ui,(t−DTi) = 0.

Figure 55 Illustration of Minimum down time (a)Valid (b)Invalid

In plot shown in Figure 55(a), the unit does not turns ON once in the window. Hence
no violation of minimum Down time. In the plot shown in Figure 55(b), the unit turns

ON twice in the window. The downtime between the two turn ONs is less than DTi.
Hence violation of minimum Down time.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 87

Crew Constraint

Typically, the number of crew in a power plant will be just sufficient to start one unit

at a time. This constraint ensures a minimum time gap between startup of two units in
the plant. Eqn. (6.16) enforces this constraint.
𝑡
∑ (∑𝑖∈𝑝 𝑠𝑖,𝑘 ) ≤ 1 (6.16)
𝑘=𝑡−𝑇𝑝

The LHS in Eqn. (6.16) is the number of unit startups in the plant in the preceding Tp

time window. This constraint can be graphically as shown in Figure 56. Units i and j
belonging to plant p honour crew constraint.

Figure 56 Illustration of crew constraint

5.7 Compensation for heat rate degradation

Presently in the calculation of heat rate degradation the entitlement of SCED in the
ISGS is taken as zero. Post SCUC this will have to be modified to the general formula

SCED Entitlement = SCED Pmax − WBES Pmax

SCED implementation is a special case wherein SCED Pmax = WBES Pmax. This leads
to 3 possibilities:

a) Zero entitlement like present when SCUC and RSD procedure provide the same
UC

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 88

 b) Positive entitlement when SCUC keeps more units onbar compared to RSD
procedure

c) Negative entitlement when SCUC keeps less units onbar compared to RSD
procedure. Constituents share compensation for heat rate degradation if

RLDC Sch < 0.85 × WBES Pmax

SCED Pool share compensation for heat rate degradation if

SCED Sch < 0.85 × SCED Entitlement

Note: These entitlements are to be used only for the purpose of heat rate

degradation calculations and not in the recovery of fixed charges.

5.8 Startup Costs due to SCUC

Startup cost (other than revival from forced outages/ planned maintenance) is
proposed to be paid from SCED pool.

5.9 Settlement Related Issues under SCUC

SCUC would imply that unit commitment would be as per merit order ensuring

sufficient spinning reserves in the system (the problem could be formulated with
region-wise reserves too).

As far as the beneficiaries are concerned, they would have the freedom to schedule as

per their entitlement based on total DC. This could lead to a situation where NR
beneficiaries have scheduled from NR plants and SCUC has de-committed some of the

NR plants. Instead, SCUC has committed some plants in SR (against which SR

beneficiaries have given nil requisition). This would in effect mean that the demand of

NR states is partly met from these plants in SR. This is something that is already
occurring under SCED albeit to a lesser extent and only for the units which are running.

But SCUC would change this to a great extent, particularly as more RE gets added in
the system.

Sufficient checks and balances would be required to ensure continuous running of

SCUC. In the above example, SR beneficiaries must ensure that they observe discipline
and don’t suddenly schedule from the SCUC committed SR station only during peak

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 89

hours. To start with, we could have a stipulation that ratio of off-peak to peak
requisitions cannot be less than 30-35%.

The Government of India has already come out separately with RE/non-RE bundling
schemes and the SCUC formulation & settlement should lead to a similar arrangement

but with significant cost savings for the parties involved and well as ensuring adequate
reserves at the inter-state level for security. This in no way absolves the beneficiaries

from maintaining similar reserves at the intra-state level too.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 90

 6 WAY FORWARD
6.1 Extension of the SCED Pilot Project
The implementation of a Pilot on SCED in ISGS Pan India was directed by Central
Electricity Regulatory Commission i.e. Central Commission w.e.f 1st April, 2019 for six

months which was extended upto 31st March, 2020 and further upto 31st May, 2020.
Central Commission, on 18 April, 2020, directed the extension and expansion of the

pilot on SCED till 31st March, 2021.

In view of Covid-19 pandemic and other associated challenges, the participation

of the partly tied/merchant/intra-state generators in the expanded SCED pilot post

June, 2020 has been minimal. Therefore, with the picking up of demand and

normalization processes, it is felt that one year extension i.e. upto 31st March, 2022 for
the SCED pilot project may be granted by the Central Commission. It would give some

space for the interested generators to participate in the pilot project. The innovation
of Look Ahead SCED or multi-period optimization can also be implemented during

this period as an improvement to SCED algorithm.

6.2 Need for SCUC pilot project

Presently, SCED optimisation is carried out after the unit commitment has taken
place at the day-ahead level for the minimisation of the variable cost or fuel charges.

Unit commitment, at ISGS level, is being handled as per the operating procedure
approved by CERC for taking units under reserve shut down. A case study on SCUC

optimization has been attempted for eight months i.e. from 01 June, 2020 to 31
January, 2021. Each run was for 3-day window on a rolling basis. The objective to

minimize the operation cost for the time window under consideration. The simulation
results have been encouraging with increase in the savings of total production costs

and scope for improved grid security in the form of reserves. The changes to the
existing scheduling activities and timelines for SCUC have been detailed in the report.

Some changes would be needed in WBES for certain aspects like Unit DC, Pmax, Pmin,
Requisitions by Constituents, Unit Commitment/De-Commitment etc.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 91

 The Central Commission constituted Expert Group to review Indian Electricity
Grid Code proposed that the Security Constrained Unit Commitment (SCUC) exercise

may be carried out to facilitate reliability of supply to the regional entities/beneficiaries

taking into account optimal cost, adequate reserves, ramping requirements factoring

security constraints. In order to ensure availability of adequate secondary and tertiary

reserves with sufficient ramping capability, NLDC may identify the generating unit for

purpose of unit commitment at the national level on a 3-day rolling basis. NLDC,
through RLDC may advise the regional entity generators to commit or de-commit the

unit.

Therefore, it is proposed that NLDC, in coordination with RLDCs, may perform

the SCUC on pilot basis for one year at the National level, for regional entity generating

stations, in order to meet total power requisitioned by regional entities/beneficiaries

at optimal cost, maintaining adequate spinning and ramping reserves subject to

security constraints as per applicable CERC regulations.

6.3 Increase Frequency of Declaration of Variable Charges

At present, the variable charges declared by the generators for the purpose of
Reserve Regulation Ancillary Services (RRAS) are considered in the optimisation

process. There is monthly declaration of variable charges by the generators under the
RRAS as per the bill raised by the RRAS provider for the generating stations or in case

that is not available, the, last available month bill. The accrual of savings due to the
optimization largely depends on spread of variable charges. However, with the optimal

fuel supply and rationalization of coal stock pan-India, the spread of variable charges
is shrinking on month-on-month basis. The reduction of spread of variable charges

may also be one of the causes for lowering of SCED savings.

It is proposed that instead of monthly declaration, the variable charges may be

declared on fortnightly basis or even weekly basis.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 92

 6.4 Declaration of Incremental Heat Rate Curves
The declaration of the variable charge is on the basis of normative PLF of 85%.

The operation below the normative PLF would entail heat rate degradation
compensation in accordance with PPA provisions and Grid Code.

Therefore, there is a need for incremental heat rate curve for every generator.
Unit input-output curves and valve point loading effects can be considered in the

optimization process as and when they are made available for every plant. Coupling
the heat rate information and granular emission data would further enhance the

optimization process.

6.5 Need to bring down Minimum Turn-Down Level

This aspect has been highlighted in the previous (January 2020) feedback report
too. The minimum turn-down level refers to the operation of despatchable generators

at low levels. It is observed that during periods of low demand, there is a need for
generators that can turn down output to low levels but remain available to rise again

quickly. Technical minimum is one of the constraints in SCED and its dual is generated
whenever the constraint is binding for any plant. Annexure-7 shows the percentage of

time various constraints were binding for different plants in 2019 and 2020.

CEA Technical Standards for Construction of Electric Plants and Electric Lines,
2010 defines ‘control load’ for coal or lignite based thermal generating units which

means the lowest load at which the rated steam temperature can be maintained under
auto control system. This is generally taken as 50% of the Maximum Continuous Rating

(MCR). CERC has mandated 55% technical minimum for all inter-state generating
stations. Further, CERC has allowed compensation for degradation of heat rate and

auxiliary consumption for operation below 85% of maximum capacity13.

13 CERC, Indian Electricity Grid Code, 2010 (4th Amendment)

http://www.cercind.gov.in/2016/regulation/124_1.pdf

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 93

 If the minimum turn-down levels are further improved, there would be more
scope for optimization and consequent, further reduction in costs.

6.6 Spinning (Hot) and Non-Spinning (Cold) Reserves

The Central Commission constituted an Expert Group to review Indian Electricity

Grid Code and other related issues in May, 2019. The expert group submitted the
report in January, 2020. The draft revised IEGC proposed by the expert group has the

provisions to ensure and monitor availability of reserve capacity. The quantum of

reserve capacity required to be maintained for grid security is related to credible

contingency including net error in the forecasts of demand and renewable generation.
The responsibility to provide reserve response may be shared by all Control Areas in a

distributed manner in the interest of grid security and in a participative manner so that
there is no tendency to pass on the responsibility to other entities.

A mandate for reserves in the appropriate Regulations/Grid Code is needed for

maintaining reserves and calling cold reserves into service as and when needed.

6.7 National Pool Account

The importance of the scheduling methodology from corridor wise scheduling

to net-injection/net-drawal for each region is again highlighted. This would require

amendments to the ‘Scheduling and Despatch Code’ under the Indian Electricity Grid
Code (IEGC). With the implementation of net-injection/net-drawal based scheduling

for each region, a ‘National Deviation Settlement (DSM) Pool’ is needed to be

maintained and operated by NLDC as a Regulatory Pool Account. The respective

Regional DSM pools may interact with the National DSM Pool only.

6.8 Statement of Station Heat Rate Degradation Compensation

NLDC has been mandated to issue the monthly Statement for heat rate
degradation compensation on account of SCED on monthly basis as per the format

SCED4 of “Procedure for Pilot on Security Constrained Economic Despatch for Inter
State Generating Stations pan India”. Central Commission has stressed on the

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 94

importance of issuance of all account statements in timely manner to facilitate smooth
disbursement and settlement.

In absence of Statement for heat rate degradation compensation on account of

SCED due to delay in issuance of the same by different RPCs, there is an impact on the

timely preparation of the “National net SCED Benefits Distribution Statement” for SCED
Generator and their respective Beneficiary.

6.9 Replication of SCED at Intra State level through SLDCs

There is a minimum basic set of data required for economic despatch for every

time block for each power station whether at inter-state level or intra-state level. The
following aspects learnt during the implementation at inter-state level may be

customized at intra-state level as follows:

 Robust Scheduling, Metering, Accounting and Settlement (SAMAST) and Intra-

State Reserves and Ancillary Services (SANTULAN) frameworks

 Visibility of Merit order stack to the operators in real time

 Regulatory norms for technical minimum and compensation

 Resource level Ramp-up / Ramp down constraints

 Scheduling and settlement of reserves through schedule based payments (vis-à-vis

actual based)
 State regulatory pool accounts to be self-sustainable

 Scheduling and Settlement of ancillary services

 IT, communication infrastructure and skilled human resources

6.10 Optimization considering full transmission network

Efforts are also underway in POSOCO to understand the impact of congestion
due to transmission on optimization of various generation resources across India.

Towards this end, simulations have been conducted using Optimal Power Flow (OPF)
technique on All India power flow model and sample locational marginal prices (LMP)

have been extracted. A summary of these efforts has been detailed at Annexure 8.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 95

 ANNEXURES

A1 Feedback to CERC on Increasing Participation in SCED Page 97

A2 Draft Ramp Handling logic Page 103

A3 In-house RRAS Software Features Page 105

A4 Equations and Formulae in RRAS & SCED Page 115

A5 Literature Survey: Security Constrained Unit Commitment Page 124

A6 Overview of SCUC Reporting Software Page 142

A7 Analysis of duals – percentage of time constraints binding in SCED Page 147

A8 Computation of Location Marginal Prices (LMP) for Indian Grid Page 149

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 96

 ANNEXURE-1

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 97

 Feedback on Increasing Participation in the SCED Pilot
(In accordance with CERC Order 08/SM/2020 dated 18th April 2020)

1. Background
Honourable commission vide Order in petition 08/SM/2020 (Suo-Motu) dated 18th of April, 2020 has extended
the Pilot on Security Constrained Economic Dispatch (SCED) from 01st June, 2020 up to 31st March, 2021. The
Commission has also expanded the scope to all the generating stations who are willing to participate in the
pilot. The following types of generators are allowed to participate in the expanded SCED Pilot:

(a) Regional entity generators

(i) Power station with part capacity tied (whose PPAs have been approved by the Appropriate
Commission),
(ii) Power station with no tied capacities (merchant generators)
(b) State generator
(i) Owned by the State entity.
(ii) State-embedded generating stations

State generators whose scheduling, metering, accounting and settlement is in place and whose scheduling
related information exchange can be enabled through the SLDC interfacing with the concerned RLDC/
NLDC.

The total capacity under the ambit of SCED as on 31st May 2020 was 53,300 MW.

2. Stakeholder Meetings for inclusion of Regional entity generators (IPPs)

After the directions of Hon'ble CERC, list of probable regional entity generators who could potentially join
SCED, was prepared. A detailed list of Regional Entity IPPs (thermal generators) who are potential candidates
for joining SCED is enclosed at Annex – 1.
Many generators at regional level have expressed interest in participating in the SCED Pilot. Accordingly, E-
meetings were conducted with stakeholders in Southern Region (6th July 2020) and Western Region (8th July
2020) to give an overview of the SCED mechanism so that the generator share better informed about various
aspects and requirements for participation in SCED. These meetings were widely attended by the regional
entity generators, regional power committees, RLDCs and NLDC and extensive deliberations were held.
Presentation given in the meeting is attached as Annex-2

The meeting in Southern Region was attended by NLC, SEIL, MEPL, IL&FS and NTPC.
The meeting in Western Region was attended by MB power, LANCO, Reliance Energy Ltd, Essar Power, Adani
Power, ACBIL, Sasan Power, Dhariwal, CGPL and Jhabua Power Limited. Subsequently, based on the
deliberations held in the meeting, Jhabua Power Limited expressed their willingness to participate in SCED
mechanism.

3. Capacity added under Expanded SCED pilot

Farakka-III (500 MW) and Korba-III (500 MW), both of which are NTPC owned stations having part untied
capacity, have joined SCED with effect from 16th June 2020. Korba-III and Farakka-III were part of the SCED
Pilot till 26th September 2019 and 30th November 2019 respectively, after which they were excluded as the
PPA for untied capacity had expired.
Lara unit 2 (800 MW) was commissioned in July 2020 which joined the first unit already under SCED.
Jhabua (600 MW) was added under SCED with effect from 17th August 2020 after ensuring all necessary
infrastructure and processes are in place. As on 25th August 2020, the total capacity under SCED is 55,700
MW.

4. Details about Jhabua Power Limited

Jhabua Power Limited is a regional entity generating station and a subsidiary of Avantha Power &
Infrastructure Limited. It is an Independent Power Producer (IPP) based at Seoni district in the State of

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 98

Madhya Pradesh. It has installed capacity of 1x600 MW. The total tied up capacity under LTA is 425 MW,
MTOA is 107 MW and merchant capacity is 68 MW. The Technical and commercial details submitted by
Jhabua Power Limited is enclosed as Annex-3. The details of LTA/MTOA are given below.

Contracted
Type of Duration
PPA Name of Beneficiary Quantum
Agreement / Year
(MW)
Long Term M.P.Power Management
20 180
PPA-1 Agreement (LTA) Company Limited
Long Term
20 Govt. of M.P. 30
PPA-2 Agreement (LTA)
Long Term Kerala State Electricity Board
25 115
PPA-1 Agreement (LTA) Limited
Long Term Kerala State Electricity Board
25 100
PPA-2 Agreement (LTA) Limited
PTC- West Bengal State
Medium Term 3 Electricity Distribution Company 107
PPA-1 Contract (MTOA) Limited
Total tiedup capacity 532
Un-tiedup capacity 68
Total 600

Further detailed deliberations for including Jhabua in SCED were held through an E-meeting on 13th August
2020 to discuss the operational and commercial aspects. The meeting was attended by Jhabua Power
Limited, WRPC, WRLDC and NLDC. The gist of discussions held is enclosed as Annex-4. During the meeting it
was decided that
a. Jhabua Power Limited would be included in SCED from 17th August 2020
b. In order to enable Jhabua Power Limited to submit the information needed for SCED in Web Based
Energy Scheduling software (WBES) of WRLDC, necessary changes were to be implemented by 14th
August 2020 including an application programming interface software for interfacing with the SCED
software at NLDC.
c. In order to get acquainted with the process of submitting the Pmax, Pmin, Ramp Up, Ramp Down
details in WBES, mock data submission was to be carried out by Jhabua Power Limited on 15th and
16th August 2020.
d. The required changes/modification in SCED software (In-house developed by POSOCO) to
accommodate Jhabua were to be carried out on 16th August,2020.

CERC Order mentions that heat rate compensation would be applicable if it is envisaged under the respective
PPA. In the case of Jhabua, WRPC informed that the heat rate compensation provisions are available in the
PPA with Kerala only and thus, would be applicable accordingly. A communication from Jhabua Power in this
regard is enclosed at Annex-5. During various discussions, WRPC confirmed that heat rate compensation for
the station would be worked out.

Consequent to the detailed deliberations, Jhabua Power Limited was included in SCED with effect from 13:15
Hrs of 17th August 2020, becoming the first IPP having part merchant quantum.

Sample plot of the pre-SCED Schedule and Optimal Schedule for Jhabua are shown in Figure-1 below.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 99

 Figure 1: Pre-SCED Schedule and Optimal Schedule
A plot of the System Marginal Price (SMP) during the same period is shown in Figure-2 below.

Figure 2:System Marginal Price (Paise/KWh)

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 100

5. Issues for further consideration
a. Declaration of Variable Charges by Generating Stations having part-tied capacity
Power plant with part merchant or full merchant capacity is allowed to participate in SCED based on the
variable charges to be declared in accordance with the methodology outlined in the CERC Order dated 18th
April 2020. The relevant clause is extracted below:
“29. Variable charge of generators other than the ISGS whose tariff is determined by the
Commission shall be considered as under:
a. For generators having full capacity tied in multiple PPAs, variable charge shall be considered as
the weighted average of the variable charges, as determined or adopted by the Appropriate
Commission.
b. For generators having only part capacity tied in PPA(s), variable charge shall be considered as the
weighted average of the variable charges, as determined or adopted by the Appropriate
Commission for the contracted portion of the capacity. This variable charge determined for the
contracted capacity shall be considered for the untied portion of the capacity as well. …”

In this regard, it is pertinent to mention that stations which are regulated, are declaring the variable charges
as per the provisions of CERC Terms and Conditions of Tariff Regulations, however in case of IPPs the variable
charges may have other component included, other than fuel charge as per the PPA signed by them with
their respective beneficiaries.
During the deliberations, Jhabua power agreed that it would provide weighted average PPA rate for SCED
and this would also be applicable for the un-tied/merchant quantum.
In addition, Jhabua stated the following:
(a) In Kerala and WB PPA, which is under Sec63 all the APC and water charges are part of Fixed Cost.
(b) PPA with MP is under Sec 62, and determined in line with TCT CERC/SERC Regulations by MPSERC and
both APC and Water charges, which are on reimbursable basis, are considered under VC.

As per the CERC Terms and Conditions of Tariff Regulations, water charges are part of O&M.

For the purpose of SCED, it was clarified to Jhabua, that the weighted average variable charges declared by
the station should be as per the billing for the previous month.

b. Treatment of capacity without having any Access in transmission system

As per the SCED extension order, power plant with part merchant or full merchant capacity is allowed to
participate under SCED. The power station participating under SCED may be required to generate up to their
maximum declared capacity (Pmax which includes both tied & untied/merchant portion) depending on the
system marginal price vis-à-vis the variable charge declared by the concerned generating station. Power
station with part merchant or full merchant capacity may be required to inject in grid including merchant
quantum which was left after all opportunities for sale of power in the market have been exhausted.

As per the Regulation 8.6 of the CERC Grant of Connectivity, Long-term Access and Medium-term Open
Access in inter-State Transmission and related matters Regulations, in order to inject power in grid, it is
mandatory to obtain some form of access i.e., long-term, medium-term or short-term. The relevant extract
is given below.
“The grant of connectivity shall not entitle an applicant to interchange any power with the grid
unless it obtains long-term access, medium-term open access or short term open access”

Further, the Regional Transmission Account (RTA), prepared in accordance with the extant POC Regulations,
considers LTA+MTOA+STOA quantum as the approved injection limit. If SCED schedules are given over and
above the LTA+MTOA+STOA quantum, then the generator would be adversely affected commercially.

Similar issues may arise in case of State Generator participating under SCED. They may not have access to
inject in ISTS network, but they may get injection schedule under SCED to ISTS.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 101

In order to comply with the above mentioned regulatory provisions, SCED schedules can thus be given for a
maximum quantum for which long-term, medium-term or short-term access has been obtained by the
concerned generating station. In this regard, schedules under SCED can also be given if the plant has taken
access to a target region without specifying the buyer.

Hence, in the case of Jhabua, to start with, Pmax for SCED has been considered as the Approved
LTA+MTOA+STOA+PX+RTM quantum only.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 102

 ANNEXURE-2
Draft Ramp Handling Logic – formulated i.c.w. NTPC
Applicable to – Thermal and Gas ISGS

Glossary –

DCt – On-bar DC in MW (scheduling limit) for tth block (Declared by ISGS)

TMt – Technical minimum in MW for tth block (Calculated based on % tech min entered by
RLDC)

DecUpt – Ramp up rate in MW/block for tth block (Declared by ISGS)

DecDnt – Ramp down rate in MW/block for tth block (Declared by ISGS, with negative sign)

St – Schedule in MW for tth block (calculated by WBES)

RUpt – Ramp up rate applicable in MW/block for tth time block (Calculated by WBES before
schedule fixation)

RDnt – Ramp down rate applicable in MW/block for tth time block (Calculated by WBES
before schedule fixation, with negative sign)

Endt – Calculated operating point in MW at end of tth time block (Calculated by WBES after
fixation of tth time block schedule. To be used during schedule fixation for t+1th time block.)

Calculation of Ramp limits for tth time block (RUpt and RDnt), during fixation of St

a) Calculate operating point at end of previous block (Endt-1)

Endt-1 = St-1 + (St-1 – St-2)/2 ; subject to upper and lower limits TMt-1 < Endt-1 < DCt-1

b) Calculation of Ramp up limit (RUpt)

Check whether Endt-1 + DecUpt > DCt ?

i) If TRUE, then
𝟏 𝑫𝑪𝒕 − 𝑬𝒏𝒅𝒕−𝟏
𝑹𝑼𝒑𝒕 = 𝑬𝒏𝒅𝒕−𝟏 − 𝑺𝒕−𝟏 + ∗ (𝑫𝑪𝒕 − 𝑬𝒏𝒅𝒕−𝟏 ) ∗ (𝟐 − )
𝟐 𝑫𝒆𝒄𝑼𝒑𝒕
ii) If FALSE, then

RUpt = St-1 – St-2 + (DecUpt/2)

In both cases, the ramp limit shall be bounded by zero and the declared ramp up rate, i.e.

0 ≤ RUpt ≤ DecUpt

c) Calculation of Ramp down limit (Rdnt)

Check whether Endt-1 + DecDnt < TMt ?
i) If TRUE, then
𝟏 𝑬𝒏𝒅𝒕−𝟏 − 𝑻𝑴𝒕
𝑹𝑫𝒏𝒕 = 𝑬𝒏𝒅𝒕−𝟏 − 𝑺𝒕−𝟏 − ∗ (𝑬𝒏𝒅𝒕−𝟏 − 𝑻𝑴𝒕 ) ∗ (𝟐 + )
𝟐 𝑫𝒆𝒄𝑫𝒏𝒕
ii) If FALSE, then
RDnt = St-1 – St-2 + (DecDnt/2)

In both cases, the ramp limit shall be bounded by the declared ramp down rate and zero, i.e.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 103

 DecDnt ≤ RDnt ≤ 0

Note that in this case, DecDnt contains a negative sign and the resultant RDnt would also be
a negative (or zero) value.

d) Schedule fixation (St)

During schedule fixation, the modified ramp limits would be used instead of declared ramp
limits. All other logic pertaining to ramp would remain intact.
St-1 + RDnt ≤ St ≤ S t-1 + RUpt

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 104

 ANNEXURE - 3

In-house RRAS Software Features

A new in-house software has been developed for dispatching RRAS instructions by
NLDC Control Room to power stations, via RLDCs. The dispatch process has been
significantly automated and streamlined, in line with the new dispatch timelines, after
implementation of RTM. The salient features of the control room software are as below:

RRAS Dashboard
RRAS dashboard (Figure A) shows the essential information about the RRAS
application.

Figure A RRAS Dashboard

The different information displayed on the dashboard is as below:

Core Engine Status

RRAS core engine runs 2 time blocks ahead of the actual dispatch in the 12th minute,
just prior to SCED core engine, i.e. in block T, RRAS will get dispatched for block (T+3).
As an example, for the 49th time block (1200-1215), RRAS will run at 1127 hrs.
Appropriate indications have been given to indicate that core engine is running
properly. The time under the box indicates the time at which last core engine run has
taken place. If the last engine run time is more than 15 minutes old, immediately offline
team would be informed.

Input Schedule Status

Every time block, RRAS software has to successfully read schedule data from NLDC
WBES database. Suitable indication has been provided to indicate the last time when
data has been fetched from RLDCs WBES to NLDC WBES. This activity starts in the 10th

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 105

minute of every time block and it should get completed by 12th minute. If the last
fetched time is not in the 10th or 11th minute, or the last fetched time is more than 15
minutes old, then follow-up regarding communication issue will be initiated by offline
team at NLDC.

RRAS/SCED Output Status

After core engine run, RRAS outputs are posted in the NLDC WBES in the 12 th minute
of every block. When the RLDCs’ schedule preparation takes place, these values are
picked up from NLDC WBES automatically. RRAS (and SCED) should ideally appear in
RLDC schedules 28 minutes before the starting of the concerned block. If the
application has not been able to write the output successfully or the output is more
than 15 minutes old, action would be initiated by offline team at NLDC.

Applied vs Implemented RRAS

A chart shows the quantum of RRAS applied by the operator and actual RRAS quantum
dispatched for all 96 blocks of current day. While the applied RRAS will appear for all
blocks for which quantum has been entered, the implemented RRAS would get
updated block-by-block as RRAS is dispatched one block at a time.

Reserves

This chart shows the trend of spinning and ramping reserves in up and down direction
on all India basis for all 96 blocks. Ramping reserves (also called 15-minute reserves)
is the quantum of generation that can be ramped up or down in one block. Reserves
are calculated considering only the stations that are included in RRAS and are on-bar.

Schedule

This chart shows the trend of schedule, DC and Technical minimum of all units which
are part of SCED and RRAS stack for 96 blocks of the current day

Exclude lists

The three tabs in this section (Figure B) show the list of stations that have been
excluded from SCED, excluded from RRAS, or provided Technical minimum support
under RRAS during the current time block. Whenever any station is excluded from
RRAS dispatch, it would show up in the RRAS exclude lists.

List of recent RRAS instructions

This list (Figure C) shows the most recent blocks in which RRAS has been dispatched
along with the RRAS quantum and the plantwise RRAS dispatch in that block is
displayed.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 106

 Figure B Exclude lists

Figure C Recent RRAS Instruction List

Application of RRAS
The Apply RRAS page (Figure D) is used to apply or withdraw RRAS. User has the
flexibility to enter RRAS quantum for any number of blocks in the future. The time
blocks are colour coded as – past blocks are shaded red; current and next two blocks
are coded yellow; and T+3 block onwards are coded green. Red and yellow blocks are
locked for user input as RRAS and SCED dispatch for these blocks has already taken
place. The RRAS dispatched is indicated in the relevant column. The green blocks are
unlocked and their value can be changed at any time until 35 minutes before the start
of the block. Ramping and Spinning Reserves are also displayed for reference of the
user as to how much reserve can be dispatched under RRAS.

RRAS would get dispatched on a block-by-block basis and it would automatically

get incorporated in the injection schedules in RLDCs WBES without any manual
intervention by RLDC shift engineers.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 107

 Current time
block

Current+3rd
time block
(editable)

Figure D Apply RRAS Interface

To apply RRAS:

Shift engineers can enter the desired MW quantum in the RRAS applied column
against the desired time block. They also record the reason (from the drop down).
Alternatively, shift engineers can select from block and to block from the drop down.
Feature was also provided to apply RRAS for next day (usually decided after 2200 hrs).

To withdraw/modify RRAS:

A feature has also been provided to withdraw already applied RRAS, where the RRAS
quantum in the particular block will be made zero.

The cells corresponding to any particular block are only open for user input until before
dispatch engine runs for that block. The data input would be locked 35 minutes before
start of the block and any actions to apply/withdraw/modify RRAS need to be done
before that. After any change in the page, Apply/Withdraw regulation button must be
clicked to save the changes.

In the old RRAS software, the RRAS dispatched was zero for the first block of dispatch.
However, this is not the case with the new software and RRAS would get dispatched
from the block in which MW quantum has been entered.

Exclusion Interface
To exclude any generator from RRAS, user needs to tick the check box against the
generator name. To exclude/include all generators of a particular region at once, the
check boxes at the top can be used. The exclusion will become effective from the next
run of RRAS. For example, if a generator is excluded at 12:05 hrs, it will be excluded
from RRAS from 52nd block onwards.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 108

 To include any generator back in RRAS, user can uncheck the excluded box and click
on Exclude button.

Technical Minimum support through RRAS

To facilitate keeping units in operation by providing technical minimum support
through RRAS, the RRAS support module has been created (Figure F). To provide
technical minimum support to any station, shift engineers select the time range and
apply support for that duration. Whenever the station’s schedule falls below technical
minimum, an equivalent amount of RRAS up would be given to ensure that the
schedule remains at 55%. All RRAS support instructions given can be withdrawn if
needed.

Figure E RRAS Exclusion Interface

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 109

 Figure F RRAS Support Interface

Modeling of Gas stations

Multi-fuel gas-based stations present a unique modeling challenge when bringing
them under optimization. As the different fuels have different variable charges,
different fuels are modeled as separate stations in the scheduling, accounting and
settlement framework. As the different fuels share the same physical equipment, the
combined generation of all fuel-types at a plant needs to be considered. This assumes
particular importance when framing the technical minimum constraint, as some
regions have the practice of not enforcing technical minimum in schedules in line with
decision taken in RPCs. In the new RRAS software, plant-level constraints have also
been framed for gas stations, in addition to fuel-level constraints. Maximum and
minimum generation constraints are calculated at plant level and it is ensured that the
total schedule with RRAS (of all fuel types) remains within this band.

Another complication in treatment of gas stations arises from open cycle/combined

cycle operation. Under present scheduling practices, DC is available for combined and
open cycle separately but the injection schedule is not segregated between both
heads. The variable charges under O/C are generally 1.5 times that under C/C, which
distorts the merit order. The upfront declaration of variable charges by stations is done
at combined cycle rates. Therefore, to avoid post-facto accounting and settlement
issues, by default, only combined cycle DC is considered for RRAS dispatch. In case
real-time engineers need to dispatch RRAS from a gas station under open-cycle, a
station-wise toggle has been provided in the user interface to enable RRAS dispatch
from open-cycle DC.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 110

Reports
Different types of graphical reports are available, which can be viewed over a selected
date range. The option to export in image formats, export in data formats is available.
A bar with movable handles provides the flexibility to zoom in and out on the time
scale. The different plots available are:

 Punched and dispatched RRAS quantum

 RRAS SMP (paise/kWh)

 Trend of spinning/ramping reserves with dispatched RRAS quantum

 Plant-wise and All India/regional schedule profile – DC, DC on bar, Tech Min,
Schedule, Schedule with RRAS (Original Schedule + RRAS), Optimal Schedule
(Original Schedule+RRAS+SCED)

Figure G Reports in RRAS software

Certainty of RRAS dispatch

One of the key benefits from implementation of revamped RRAS software has
been the increased certainty of getting the desired RRAS dispatch quantum. With the
previous implementation, due to revisions by constituents, dispatched RRAS over a
longer time duration used to carry uncertainty. With the new scheduling timelines and
single-period RRAS dispatch (at a time), the quantum of RRAS identified by the engine

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 111

is almost certain to get implemented in the net injection schedule. This aspect can be
seen in Fig where the implemented RRAS quantum with the old application (Dec 19)
appears smooth owing to moderation and curtailment at WBES end due to
constituents’ revisions. In contrast, with the new application (Dec 2020), the
implemented RRAS is much closer to the desired RRAS quantum and thus appears like
a stepped curve (corresponding to RRAS quantum required by real-time engineers).

Since the implementation of the new software, the average RRAS performance
index (RRAS dispatched/RRAS desired) has been 93.4%. This can be seen in terms of a
duration curve in Figure K.

Figure H Duration curve of implemented RRAS: Dec 2019 vs Dec 2020

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 112

 Daily Energy Dispatch in RRAS (Jan 2019 onwards)
30 RRAS Up RRAS Down

20

10
RRAS Dispatched MU)

-10

-20

-30

-40

-50

Figure I Daily RRAS Up and Down dispatch energy (MU)

Daily RRAS Dispatch Comparison - 2019 & 2020

40 Up-2019 Down-2019 Up-2020 Down-2020

30

20

10
RRAS Dispatched (MU)

-10

-20

-30

-40

-50
01-Jan 01-Feb 01-Mar 01-Apr 01-May 01-Jun 01-Jul 01-Aug 01-Sep 01-Oct 01-Nov 01-Dec

Figure J Daily RRAS Dispatch Comparison – 2019 & 2020

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 113

 Figure K Duration of RRAS dispatched & implemented (since 19 Sep 2020)

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 114

 ANNEXURE - 4
Equations and Formulae in RRAS and SCED1
RRAS: Reserve Regulation Ancillary Services

SCED: Security Constrained Economic Dispatch

g is the superset of power plants. Participation in SCED and RRAS handled through respective
Exclude flags. Different fuel types in multi-fuel gas plants are modelled independently in ‘g’

Qg is the subset of RRAS generators from the set ‘g’. It is a dynamic set and its elements are decided
based on the RRAS include flag read by the program.

r is the set of regions: N, W, S, E, NE also identified by 1,2,3,4,5 respectively

Gas is the set of gas power plants incorporating all fuel types in a power station

DC g is the total Declared Capability of power plant g

DConbarg is the On-bar Declared Capability of power plant g

Normbarg is the Capacity on bar less normative auxiliary consumption of power plant g

TM g is the Technical Minimum considered in the scheduling of power plant g

Pmaxg  max(min( DCg , DConbarg ), Scheduleg )

Pming  min(max(0.55* Normbarg , TM g ), Scheduleg )

If Techmin Support is ON, Pming  max(0.55* Normbarg , TM g )

SCHIRr is the scheduled import of region r before SCED. (import is positive; export is negative)

ImpATCr is the import ATC of region r . Always positive.

ExpATCr is the export ATC of region r . Always positive.

Handling Exclusion:

Exfrrasg is the flag indicating exclusion status of generator from RRAS, 0 - exclude or 1 – include;
obtained from user interface

Exfsced g is the flag indicating exclusion status of generator from SCED, 0 - exclude or 1 – include;
obtained from user interface

Exfrrasg  0 , if Emergency Stop is activated

Exfsced g  0 , if Emergency Stop is activated

1
As this formulation has been highly customized for the evolving Indian power system market structure, it is
under a continuous refinement and revision process based on actual operational use-cases and hence liable to
change over time.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 115

Gas generator special considerations

DConbarCCg is the On bar Combined Cycle Declared Capability of Gas power plant g

DConbarOCg is the On bar Open Cycle Declared Capability of Gas power plant g

For gas generator fuel types, DConbarg  DConbarCCg  DConbarOCg

For gas generator fuel types, Pmaxg  max(min( DCg , DConbarg )  DConbarOCg , Scheduleg )

For gas generator fuel types, Pming  min(TM g , Scheduleg )

h is the individual identifier for each physical Gas station in g

For each Gas power plant, PmaxGas ,h   Pmax

gh
g

For each Gas power plant, NormbarGas ,h   Normbar

gh
g

For each Gas power plant, PminGas ,h  min(max(0.55* NormbarGas ,h ,  Pmin ),  Schedule )
gh
g
gh
g

For each Gas power plant, if Techmin Support is ON, then

PminGas ,h  max(0.55* NormbarGas ,h ,  Pming )
gh

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 116

 RRAS Program
Creating RRAS subset: If Exfrrasg  1, Qg  g

For all ‘g’ whose Techmin Support is ON, RRASsuportQg  max(( PminQg  ScheduleQg ),0)

ScheduleQg  ScheduleQg  RRASsuportQg . This would result in extending the techmin support as
a part of Schedule.

PrevscheduleQg is the previous time block schedule without RRAS & SCED in each generator g

PoldoptschedQg is the previous time block optimal schedule with RRAS & SCED

PrevRRAS g is the previous time block implemented RRAS in each generator g

ResUPavlblQg  min(( PmaxQg  PoldoptschedQg ), RampUpQg ) is the incremental up reserve

dispatchable in the system w.r.t. previous block optimal schedule as basis in each generator

ResDNavlblQg  min(( PoldoptschedQg  PminQg ), RampDNQg ) is the incremental down

dispatchable reserve in the system w.r.t. previous block optimal schedule as basis in each generator

PresUPavlbl  max(( ResUPavlblQg  ( ScheduleQg   PrevscheduleQg )),0)

Qg Qg Qg

PresDNavlbl  max(( ResDNavlblQg  ( ScheduleQg   PrevscheduleQg )), 0)

Qg Qg Qg

PoldRRASimpl   PrevRRASQg , is the total RRAS implemented in the previous time block
Qg

During partial exclusion of gas fuel stations from RRAS, Gas power plant limits are recalculated as

PmaxGas ,h ,Qg = PmaxGas ,h   Scheduleg -  Schedule Qg

gh Qg h

PminGas ,h ,Qg = PminGas ,h   Scheduleg -  Schedule Qg

gh Qg h

For Gas Stations which are provided Technical Minimum Support through RRAS

g  h
if Technical minimum support is provided for that plant Gash
then pTM sup GasQg  1

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 117

Provision for ramping down RRAS quantum for generators excluded from SCED

DefaultRRAS g is the RRAS quantum required to satisfy ramp for generators included in RRAS but
excluded in SCED.

If Exfrrasg  1 and Exfsced g  0 ; then OldRRASimplgeng  PrevRRASQg

 
If Exfsced g  0 and Exfrrasg  1 and OldRRASimplgeng  0 ; then DefaultRRAS g  0


If Exfsced g  0 and Exfrrasg  1 and OldRRASimplgeng  0 ; then

DefaultRRAS g  max OldRRASimplgeng  Rampdowng   Scheduleg  Prevscheduleg  , 0 

If Exfsced g  0 and Exfrrasg  1 and OldRRASimplgeng  0 ; then

DefaultRRAS g  min OldRRASimplgeng + Rampupg   Scheduleg  Prevscheduleg  , 0 
ScheduleQg  ScheduleQg  DefaultRRASQg , to incorporate this RRAS quantum in schedule

**********************************************************************************

Calculation of ‘ramp limited and schedule adjusted’ spinning reserve

RRASpunched is the RRAS punched by the shift engineer from software user interface

RRASdelta is the incremental RRAS required in this time block considering already implemented
RRAS in the previous time block

 
If RRASpunched  0 ; then RRASpunched  max  ( RRASpunched 
  DefaultRRAS Qg ), 0 

 Qg 

 
If RRASpunched  0 ; then RRASpunched  min  ( RRASpunched 
  DefaultRRASQg ), 0 
 Qg 
RRASdelta  RRASpunched  OldRRASimpl

RRASdelrestri is the restricted incremental RRAS considering the available reserve

RRASdelrestri  RRASdelta

If  RRASdelta  0 and RRASdelta  ResUPavlbl  then RRASdelrestri = ResUPavlbl

If  RRASdelta  0 and RRASdelta   Re sDNavlbl  then RRASdelrestri =  Re sDNavlbl

**********************************************************************************

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 118

RRAS Variables
i. Variables which can be positive, negative or zero
vCostrras : Objective function cost of RRAS to ensure merit order despatch

vRRASQg : Quantum of RRAS MW dispatched per generator

ii. Positive Variables

vArtrrasupQg : Artificial variable inserted in RRAS Ramp UP equation

vArtrrasdownQg : Artificial variable inserted in RRAS Ramp DOWN equation

vArtrrasposQg : Artificial variable inserted in RRAS supply balance equality constraint

vArtrrasmaxQg : Artificial variable inserted in max gen after RRAS dispatch constraint

vArtrrasminQg : Artificial variable inserted in min gen after RRAS dispatch constraint

vArtGasmaxrrasGas : Artificial variable inserted in the Gas plant level max gen constraint

vArtGasminrrasGas : Artificial variable inserted in the Gas plant level min gen constraint

iii. Negative Variables

vArtrrasnegQg : Artificial variable inserted in RRAS supply balance equality constraint

RRAS Constraint Equations

eSumRRAS total RRAS applied equality constraint
eWrrasmaxQg max limit of generation after RRAS
eWrrasminQg min limit of generation after RRAS
eRUPrrasQg ramp up limit on the movement of generator
eRDNrrasQg ramp down limit on the movement of generator
eobjfnrras objective function RRAS
eUnidirposQg RRAS direction and generator direction be same
eUnidirnegQg RRAS direction and generator direction be same
eUnidirzeroQg RRAS zero means zero dispatch
eGasMaxQGas ,Qg Plant level Max check for Gas
eGasMinQGas ,Qg Plant level Min check for Gas
eObjfnrras Objective function to minimize total cost with RRAS

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 119

eSumRRAS ..  RRASQg  RRASdelrestri  OldRRASimpl   vArtrraspos Qg +  vArtrrasneg Qg
Qg Qg Qg

eWrrasmaxQg .. ScheduleQg  vRRASQg  PmaxQg  vArtrrasmaxQg

eWrrasminQg .. ScheduleQg  vRRASQg  PminQg  vArtrrasminQg
eRUPrrasQg .. if   RRASdelrestri  OldRRASimpl   0  then
ScheduleQg  vRRASQg  OldoptschedQg  RampupQg  vArtrrasupQg
eRDNrrasQg .. if   RRASdelrestri  OldRRASimpl   0  then
ScheduleQg  vRRASQg  OldoptschedQg  RampdownQg  vArtrrasdnQg
eUnidirposQg .. if   RRASdelrestri  OldRRASimpl   0  then RRASQg  0

eUnidirnegQg .. if  RRASdelrestri  OldRRASimpl   0 and pTMGas < 1  then RRAS Qg Qg 0

eUnidirzeroQg .. if   pRRASdelrestri  pOldRRASimpl   0 and pTMGas < 1 then RRAS Qg Qg 0

eGasMaxGas ,Qg ..  Schedule
Qg h
Qg   RRAS
Qg h
Qg  PmaxGas ,Qg  vArtGasmaxrrasGas ,Qg

eGasMinGas ,Qg ..  Schedule

Qg h
Qg   RRAS
Qg h
Qg  PminGas ,Qg - vArtGasminrrasGas ,Qg

RRAS Objective Function

Minimize vCostrras (Cost of RRAS Dispatch)

vCostrras =
( RRASQg *VCQg   M *1.05* vArtrra supQg   M *1.05* vArtrra s dnQg 
Qg Qg Qg

 M *1.1* vArtrraspos
Qg
Qg   M *1.1* vArtrrasnegQg 
Qg

 M *1.1* vArtrrasmax
Qg
Qg   M *1.1* vArtrrasminQg 
Qg

 M *1.2* vArtGasmaxrras
Gas
Qg   M *1.8* vArtGasminrrasQg ) *10 /100000
Gas

Mimic insertion of RRAS into WBES Schedule for creating net schedule as below

g , RRAS g  RRASQg  RRASsupportQg  DefaultRRASQg

Schedulewrras g = Schedule g + RRASg

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 120

 SCED Program
Acting on the SCED exclusion flag & calculating limits

If Exfsced g  0, then Pmaxg  SchedulewRRAS g

If Exfsced g  0, then Pming  SchedulewRRAS g

Scheduled Stop / Soft Landing logic; Soft Landing not done when emergency stopped

If Exfsced g  0 and EmerStopg  1,

 
then Pmaxg  Oldoptsched g  min Rampupg , max  Rampdowng ,   Oldoptsched g  Schedulewrrasg  
If Exfsced g  0 and EmerStopg  1,

 
then Pming  Oldoptsched g  min Rampupg , max  Rampdowng ,   Oldoptsched g  Schedulewrrasg  
h is the individual identifier for each physical Gas station in g

For each Gas power plant, PmaxGas ,h   Pmax

gh
g

For each Gas power plant, NormbarGas ,h   Normbar

gh
g

For each Gas power plant, PminGas ,h  min(max(0.55* NormbarGas ,h ,  Pmin ),  Schedule )
gh
g
gh
g

For each Gas power plant, if Techmin Support is ON, then

PminGas ,h  max(0.55* NormbarGas ,h ,  Pming )
gh

SCED Variables
i. Variables which can be positive, negative or zero
vCost : Objective function cost of SCED

ii. Positive Variables

vGeng : Optimal Schedule of each generator after SCED

vArtmaxg : Artificial variable inserted in the max gen after SCED constraint

vArtming : Artificial variable inserted in the min gen after SCED cosntraint

vArtimpATCr : Artificial variable inserted in region wise net import ATC equation of SCED

vArtexpATCr : Artificial variable inserted in region wise net export ATC equation of SCED

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 121

 vArtScheduleg : Artificial variable inserted in the SCED Demand Supply equality constraint

vArtRampupg : Artificial variable inserted in the Ramp UP constraint of SCED

vArtRampdowng : Artificial variable inserted in the Ramp DOWN constraint of SCED

vArtGasmaxGas : Artificial variable inserted in the Gas plant level max gen constraint

vArtGasminGas : Artificial variable inserted in the Gas plant level min gen constraint

iii. Negative Variables

vArtschedming : Artificial variable inserted in the SCED Demand Supply equality constraint

SCED Constraint Equations

eATCimportr Import ATC limit constraint
eATCexportr Export ATC limit constraint
eSchedule SCED equality Constraint for demand supply balance
eMaxgeng Maximum generation limit constraint
eMingeng Minimum generation limit constraint
eRampuplimg Ramp up limit on movement of generators
eRampdnlimg Ramp down limit on movement of generators
eGasMaxGas Plant level Max check for Gas
eGasMinGas Plant level Min check for Gas
eObjfn Cost of generation to be minimized

 
eATCimportr .. SCHIR r    vGeng -  Scheduleg   min(ImpOpATCr , ImpATCr ) + vArtimpATCr
 gr gr 
 
eATCexportr .. SCHIR r    vGeng -  Scheduleg   -min(ExpOpATCr ,ExpATCr ) - vArtexpATCr
 gr gr 
eSchedule..  vGeng =  SchedulewRRAS   vArtScheduleg   vArtSchedming
g g g g

eMaxgeng .. vGen g  Pmax g + vArtmax g

eMingeng .. vGen g  Pming - vArtmin g
eRampuplimg .. vGen g  Oldoptsched g + Rampup g + vArtRampup g
eRampdnlimg .. vGen g  Oldoptsched g - Rampdown g - vArtRampdown g
eGasMaxGas,h ..  vGen
gh
g  PmaxGas + vArtGasmax Gas

eGasMinGas,h ..  vGen
gh
g  PminGas - vArtGasmin Gas

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 122

SCED Objective Function
Minimize vCost (Total cost of generation)
vCost =
( vGeng *VCg +
g

 M *0.9* vArtimpATC   M *1.1* vArtSchedule

r
r
g
g 

  M *0.9* vArtexpATCr   M *1.1* vArtmaxg 

r g

  M *1.05* vArtRampupg  M *1.05* vArtRampdowng 

g g

  M *1.1* vArtming  M *1.1* vArtSchedming 

g g

  M *1.2* vArtGasmaxGas   M *1.2* vArtGasminGas ) *10 /100000

Gas Gas

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 123

 ANNEXURE - 5

Literature Survey:
Security Constrained Unit Commitment (SCUC)
The Security Constrained Unit commitment (SCUC) is a mathematical problem,
which determines how to run all the available generating units in a power system,
and meet the forecasted demand over a time-horizon with a least cost solution.
Production cost being the major cost in system-operation, the choice of running
the efficient unit has a significant impact on the economy of the power system.
The typical time period of optimization is of one day to one week. SCUC can be
broadly consists of two parts
1. Scheduling the start-up, operation, and shutdown of the available units
2. Allocating the total power demand among the available generation units by
minimizing the system costs.
The Unit commitment uses the binary variables to represent the start –up and
shut-down decisions and the cost-optimization problem is formulated as a mixed-
integer linear program (MILP).
The system costs majorly consist of the following
1. Start – Up cost & Shut – down cost: - The start-up cost represent the
utilization of the secondary fuel oil consumption. They are non-zero values
whenever the generating unit’s commitment status is changed. The cost of
ramping also if applicable will be added to problem formalization.
2. Fixed Cost :- The annual fixed cost can consist of the components of
return on equity, interest on loan capital, depreciation, interest on working
capital; and operation and maintenance expenses
3. Variable cost: This cost would be determined by fuel prices and corrected
by efficiency (for taking into consideration all levels of generating units
output). Fuel costs of a unit are generally represented by a convex quadratic
heat consumption curve as a function of generation multiplied by the fuel
price.
The main objective of Unit commitment being to meet the forecasted demand
for each period of the optimization. Besides this, there are several system
requirements/constraints which are to be honored while finalizing the unit
commitment

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 124

 1. Spinning Reserve requirement: Reserve represent the generating
capability that is available in case of an event/contingency such as generator
or transmission line outage and also demand or variable generation forecast
error. Secondary and Tertiary reserve
2. Generator Limits
a. Ramping Capability : Ramp – Up and Ramp – Down rate
b. Start – Up and Shut-down Time : The amount of time required for
starting up – unit depending on the state of the unit i.e. Cold, Warm
and Hot
i. Simultaneous Start-Ups: The number of simultaneous start-
ups of units in the same plant is generally limited, these
constraints are modelled by the time required to bring a unit
on-line and to shut down a unit
c. Maximum/Minimum Generation Levels: The maximum
generation level is maximum output limited by the available capacity,
and minimum generation level is the minimum power – output or
stable generation level if committed.
d. Minimum Up and Down Time: Due to physical characteristics of
the generators, once a unit is started it cannot be shut down
immediately and conversely if the unit is shut down it may not start
immediately.
3. Regulation Requirements: Regulations refers to the generator action
which would correct for continuous load changes, and restore the system
frequency to 50 Hz.
4. Network Related Constraints :- Transmission security constraints are
used to capture transmission limitations i.e the real power flow restriction
over a transmission line or a transmission corridor
5. Storage Constraints: Generation units with energy storage capabilities like
large hydro reservoirs and pumped hydro storage are restricted to the
amount of energy stored/generated.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 125

 Figure 1Typical Flow Chart of Unit Commitment
The typical flow chart of unit commitment is shown in Figure -1, the dispatch of
unit continues by loading the next lowest cost and stops the process until the
desired Load/Generation is met or no more units are available to be dispatched.
The generating units are dispatched within their upper and lower limits subject to
reserve requirements and all the constraints described above. A feasible dispatch
using the above methodology is the most economical dispatch.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 126

International Experience
MIDWEST ISO
The Midwest ISO is an independent, non-profit organization in 15 U.S. States
and one Canadian province with a peak demand of 127 GW and Installed
Capacity of 177GW with 460 Market participants and 6600 Generating units. The
Midwest ISO launched its competitive wholesale electricity markets on April 1,
2005. These markets include day-ahead (DA) and real-time (RT) energy markets
and a market for Financial Transmission Rights (FTRs). The Midwest ISO
launched its competitive wholesale electricity markets on April 1, 2005. These
markets include day-ahead (DA) and real-time (RT) energy markets and a market
for Financial Transmission Rights (FTRs).
Wholesale electricity markets require a set of integrated market products that
work together to efficiently price energy and ancillary services. While the energy
markets provides transparent and efficient pricing signals for economic and
reliable grid operations in general, market prices for AS capacities are missing and
the energy markets alone are unable to efficiently price electricity during hours of
scarcity
The security-constrained commitment and dispatch models implemented for
Midwest ISO’s DA co-optimized energy and ancillary service market
The Mid-west ISO Day ahead Market products, Available resources and the key
market rules are given below
1. Market Products
a. Energy Offers in $/Mwh
i. Must – Run units are self-committed and are dispatched a
minimum of self-schedule MW levels
b. Ancillary Services offers in $/MW. The ancillary services reserve is
an hourly product in the Day – Ahead market. The ancillary services
include
i. Regulating reserve
ii. Spin Reserve and Supplementary reserve(SUP)
2. The Resources participating in the Day – Ahead market are explained
below.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 127

 a. Internal generation resources – These resources provide both Energy
and Ancillary services subject to
Max/Min/Emergency/Economic/Regulation limits, Min and Max –
Up/Down times, Start-up costs and ramp-rate limits.
b. External Asynchronous Resources (EAR) – This resources is
dispatched from 0 to its max MW subject to it ramp-rate.
c. Demand Response Resources (DRR) : There are two types of DRRs
i. Type – 1: These resources provide fixed-target MW reduction
or provide contingency reserve subject to curtailment cost,
minimum interruption/non-interruption times.
ii. Type – 2: These resources have the commitment and dispatch
models as generation resources
d. Intermittent Renewable Resources (IRR): These resources are self-
committed ad only provide energy with prescribed MW profiles.
The bids and offers for energy and ancillary services are cleared for all market
resources for the whole DA market period via a simultaneously co-optimized,
MIP-based SCUC and LP-based SCED algorithms. Transmission security
constraints are enforced in the form of predefined constraint list for the SCUC
and a simultaneous feasibility testing (SFT) function iterating with SCED. The
simultaneously co-optimized SCUC algorithm determines the overall least cost
resource commitment and dispatch schedules to meet the energy supply and
demand bids, subject to transmission security constraints, the Midwest ISO AS
requirements and resource operational parameters, such as economic maximum
and minimum limits, regulation maximum and minimum limits, ramp-rate limits,
min/max-run time and min-down time.
After the resource commitments are determined by the SCUC algorithm,
the SCED algorithm is performed to calculate DA market MW awards, LMPs
and MCPs on an hourly basis for committed dispatchable/self-scheduled
generation resources, price sensitive demand bids, external bilateral transactions,
virtual supplies, virtual demands and committed demand response resources to
supply the hourly fixed demand bid.
The DA market clearing observes the transmission security
constraints under selected contingencies and satisfies the Midwest ISO’s
ancillary service requirements.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 128

 Figure 2 Mid-West ISO Day Ahead Market and Reliability Assessment Commitment

Midwest ISO DA Market Clearing

Midwest ISO day-ahead market clearing determines hourly commitment and
dispatch schedules for ~1500 resources totaling ~130,000 MW capacity and a
peak market load of 109,157 MW. The MIP based SCUC algorithm solves the
day-market commitment problem within ~15 minutes for a MIP gap of 0.05%.
The DA SCED and SFT iterative process calculates the bid and offer awards,
LMPs for ~2000 commercial pricing nodes and MCPs for 9 local reserve zones
within ~2 hours for a network model of ~36,000 buses.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 129

California ISO

The California Independent System Operator (CAISO) is a non-

profit Independent System Operator (ISO) serving California. At peak, Cal-ISO
controls 45,000 megawatts. It is the second largest grid in the U.S. and the 5th
largest in the world spread across 26,000 circuit miles of transmission lines, with
31,208 market transactions per day and 1080 power plants. The ISO operates
California’s wholesale electricity market, using state-of-the-art technology to
match demand with the lowest cost energy available at any given time. The market
runs an automated auction every day and every five minutes, across 9,700 price
nodes, generating over 31,000 transactions a day.

CAISO uses SCUC for commitment of generating units in Day Ahead Market
and Hour-Ahead Scheduling process (HASP). SCUC is used for meeting the
CAISO Demand Forecast in

a. Market Power Mitigation – Reliability Requirement Determination

(MPM-RRD):- This is the first market process in the day-ahead market.
In this process, the markets bids are mitigated due to lack of competition
and potentially the energy prices could be manipulated by controlling the
supply. This runs for the coming 24 Hours and with Hourly resolution.
This process is also evaluated in Real-Time Market.
b. Integrated Forward Market (IFM) :- The IFM is a market for trading
physical and virtual Energy and Ancillary Services for each Trading Hour
of the next Trading Day
c. Residual Unit Commitment (RUC):- Residual Unit Commitment is a
reliability function for committing resources and procuring RUC Capacity
not scheduled in the IFM (as physical Energy or AS capacity.) RUC
Capacity is procured in order to meet the difference between the CAISO
Forecast of CAISO Demand (CFCD) (including locational differences and
adjustments) and the Demand scheduled in the IFM, for each Trading
Hour of the Trading Day
d. HASP :- The Hour-Ahead Scheduling Process (HASP) is a process for
scheduling Energy and AS based on the Bids submitted into the HASP
e. Real-Time Unit Commitment (RTUC):- It is a market process for
committing fast and Short-Start Units and awarding additional AS from
Dynamic System Resources at 15-minute intervals. The RTUC function

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 130

 runs every 15 minutes and looks ahead in 15-minute intervals spanning the
current Trading Hour and next Trading Hour
f. Short Term Unit Commitment (STUC):- Short-Term Unit
Commitment (STUC) is a reliability function for committing Short and
Medium Start Units to meet the CAISO Forecast of CAISO Demand. The
STUC function is performed hourly and looks ahead three hours beyond
the Trading Hour, at 15-minute intervals

Figure 3 CAISO Day Ahead Market Time-Line

The Day-Ahead Market Clearing problem includes next-day Generation and

Demand Bids. The objective of the problem is to minimize Energy and Ancillary
Services (AS) procurement costs subject to all submitted Energy and Ancillary
Services submitted supply bids and transmission constraints. A similar
formulation is used to solve the Real-Time Market Clearing problem as well as
the Residual Unit Commitment problem
In Real-Time, unit commitment is limited to medium- and fast-start units
and the dispatch is initialized from the State Estimator solution or telemetry. The
SCUC commits and dispatches resources based on minimum cost as reflected by
Bid prices, subject to network constraints. The linearized transmission constraints
are identified using AC-based power flow and contingency analysis algorithms
based on a Full Network Model (FNM). The FNM includes all CAISO Balancing

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 131

Authority Area transmission network buses and transmission constraints, and
possibly a reduced network representation of the rest of the WECC system
SCUC calculates Locational Marginal Prices (LMPs) for Energy, network
constraint Shadow Prices and Ancillary Services Marginal Prices (ASMPs)
consistent with the AC-based power flow model. SCUC employs a Mixed Integer
Programming (MIP)

The SCUC co-optimization engine in CAISO clear’s markets for Energy and
Ancillary Services including the following modeling and functional capabilities:
1. Simultaneous optimization of the following:
a. Energy
b. Regulation Up and Down
c. Spinning and Non-Spinning Reserve
d. Reliability capacity
2. Least-cost Market Clearing based on:
a. Three-part Generation Energy Bids – Start-Up, Minimum Load and
Energy Bid cost
b. Single-part load Bids ($/hr.)
c. Single-part Inter-Tie Energy Bids
d. Ancillary Services Bids
e. RUC Availability Bids
3. Network Congestion Management
a. Full AC network model including transmission losses
b. Security analysis (contingency constraints)
c. Nomogram constraints
4. Marginal Pricing
a. Energy, network loss and transmission congestion LMP components
b. Ancillary Service prices for each Ancillary Service Region and each
Ancillary Service Bid
c. RUC Prices

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 132

New York ISO

The NYISO is the New York Independent System Operator — the organization
responsible for managing New York’s electric grid and its competitive wholesale
electric marketplace. New York has a peak demand of 33,956 MW with a total of
41,319 megawatts (MW) of power resources available

At New York ISO Security constraint Unit Commitment creates the ISO’s Day –
Ahead Market Schedules and prices by considering
1. Load Forecast
2. Ancillary Service requirement as determined by ISO
3. Bilateral Transaction schedule
4. Price Bids and Operating constraints submitted for Generator/Demand
Side resources
5. Price Bids for Ancillary Services
6. Bids to purchase or sell Energy from or to the Day-Ahead Market

The SCUC algorithm simultaneously minimizes total Bid Production Cost of

1. Committing sufficient capacity to meet the ISO’s Load forecast
2. Providing Ancillary services – Regulation and Operating Reserve

NYISO would consider Demand reduction bids, if total Bid production cost
reduction is possible. Further, NYISO considers substitute Higher Quality
Services(i.e. shorter response time) for lower quality Ancillary service when doing
so would result in overall least bid cost solution. (10 Min Non-Synchronized
reserved maybe substituted for 30-Minute Reserve)

NYISO would commit one or more Generator(s) in the Day-Ahead Market for a
Dispatch Day if it determines that the Generator(s) are needed to meet NYCA
reliability requirements.
After the Day-Ahead schedule is published, NYISO will evaluate any events like
the loss of significant Generators or transmission facilities that may cause the
Day-Ahead schedules to be inadequate to meet the Load or reliability
requirements for the Dispatch Day. In such scenarios NYISO would either
1. commit additional Resources, beyond those committed in Day-Ahead
Market and considering

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 133

 a. Bids submitted that were previously not accepted but were
designated by the bidder as continuing to be available
b. new Bids from all Suppliers, including neighboring systems
2. Provide Notice to all resources to run above their Day-Ahead Despatch
levels up to the Max Capacity levels

Reliability Forecast at NYISO

NYISO to preserve system reliability, and ensure that there will be sufficient
resources available to meet forecasted Load and reserve requirement over the
seven day period that begins with the next Dispatch Day, a Supplemental
Resource Evaluation will be performed. If it is determined that a long start-up
time Generator (i.e., a Generator that cannot be scheduled by SCUC to start up
in time for the next Dispatch Day) is needed for reliability, the ISO shall accept a
Bid from the Generator and the Generator will begin its start-up sequence.
During each day of the start-up sequence, NYISO will perform an SRE to
determine if long start-up time Generators will still be needed as previously
forecasted. If at any time it is determined that the Generator will not be needed
as previously forecasted, the ISO shall order the Generator to abort its start-up
sequence. The ISO will commit to long start-up time Generators to preserve
reliability. However, the ISO will not commit resources with long start-up times
to reduce the cost of meeting Loads that it expects to occur in days following the
next Dispatch Day.

Real-Time Commitment (“RTC”)

After the Day-Ahead schedule is published and no later than 75 minutes before
each hour, Customers may submit Real-Time Bids into RTC for real-time
evaluation. RTC will make binding unit commitment and de-commitment
decisions for the periods beginning fifteen minutes (in the case of Resources that
can respond in ten minutes) and thirty minutes (in the case of Resources that can
respond in thirty minutes) after the scheduled posting time of each RTC run. RTC
will co-optimize to solve simultaneously for all Load, Operating Reserves and
Regulation Service and to minimize the total as-bid production costs over its
optimization timeframe. RTC will consider SCUC’s Resource commitment for
the day, load forecasts that RTC itself will produce each quarter hour, binding
transmission constraints, and all Real-Time Bids and Bid parameters

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 134

Real-Time Dispatch
The Real-Time Dispatch will make dispatching decisions, send Base Point Signals
to Internal Generators and Demand Side Resources, produce schedules for intra-
hour External Transactions at Dynamically Scheduled Proxy Generator Buses,
calculate Real-Time Market clearing prices for Energy, Operating Reserves, and
Real-Time Market Prices for Regulation Service, and establish real-time schedules
for those products on a five-minute basis, starting at the beginning of each hour.
The Real-Time Dispatch will not make commitment decisions and will not
consider start-up costs in any of its dispatching or pricing decisions

Figure 4 NYISO Bid to Bill System Process

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 135

NYISO Locational Based Marginal Prices (LBMP) Timeline
The Market participants need to submit the Day-Ahead bids by 05:00 AM on the
day prior to the Dispatch Day for the full commitment period. By 11:00 AM on
the day prior to Dispatch Day, NYISCO would complete the Day-Ahead
Scheduling Unit Commitment process and post the Day-Ahead Schedule. The
LBMP are posted as public data and commitment schedule as private data. The
following functional parameters are used in the Day-Ahead SCUC
1. Weather & Load Forecast – The Load forecast function is used to
forecast load for each zone in NYCA. The load forecast function used
historical load and weather data information for each zone and develops
forecast models.
2. Reserve and Regulation Requirement : The SCUC function obtains the
following hourly requirements from EMS
a. Spinning 10 – Minute reserve
b. Total 10 – Minute reserve ( Include 10-
minute reserve)
c. Total operating reserve ( 10 – minute
reserve and 30 – minute reserve)
d. Regulation capacity
3. Generator and Transmission parameters
a. Current Generating unit operating status
b. Constraints on the minimum up and
down time of the generators
c. Generating and Start up Bid prices
d. Plant-related startup and shutdown
constraints
e. Minimum and Maximum generation
constraints
f. Transmission maintenance schedules
g. Transmission constraints
4. Outage Scheduler : NYISO uses this Figure 5 NYISO LBMP Time Line
function to keep a track of scheduled
equipment outages and NYCA and it provides a user interface for entering
equipment outage schedules, as well as reviewing existing schedules

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 136

PJM
PJM Interconnection LLC (PJM) is a regional transmission organization (RTO)
in the United States. PJM, was the world's largest competitive wholesale electricity
market with More than 1,000 companies as members of PJM, which serves 65
million customers and has 180 gigawatts of generating capacity. With 1,376
generation sources, 84,236 miles (135,560 km) of transmission lines and 6,038
transmission substations, PJM delivered 807 terawatt-hours of electricity in 2018
Scheduling by PJM includes the
a. Day Ahead Energy Market
b. Control Area reliability – based scheduling process
c. Hourly Scheduling process
PJM’s responsibilities are to support scheduling activities for all PJM Members
including:
1. Developing the Day-ahead Market financial schedules based upon
participant-supplied bids, offers and bilateral transaction schedules using
least-cost security constrained unit commitment (SCUC) and dispatch
analysis.
2. Post the following information after the Day-ahead Market clears at 4:00
PM
3. Schedules for Next Day by participant (generation & demand),
4. Transaction Schedules
5. Day-ahead LMPs , Binding Transmission Constraints, Net Tie Schedules,
Reactive 500 kV Interface Indicator Limits
6. PJM Load Forecast
7. Aggregate Demand Bids
8. PJM Operating Reserve Objective.

Perform scheduling for the Forecasted load and reserves not covered by the Day-
ahead Demand bids, Self-Scheduled Resources or Bilateral Transactions,
including scheduling generation to relieve expected transmission constraints

Clear the Regulation Market and Spinning Reserve Markets Simultaneously and
post the Regulation Marginal Clearing Price (RMCP) and Spinning Reserve
Marginal Clearing Price (SRMCP) on an hourly basis no later than 30 minutes
prior to the start of the operating hour.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 137

PJM Two Settlement system
The Two-Settlement system consists of two markets,
1. Day-ahead market and a
2. Real-time balancing market.
The Day-ahead Market is a forward market in which hourly clearing prices are
calculated for each hour of the next operating day based on generation offers,
demand bids, Increment offers, Decrement bids and bilateral transaction
schedules submitted into the Day-ahead Market. The balancing market is the real-
time energy market in which the clearing prices are calculated every five minutes
based on the actual system operations security-constrained economic dispatch.
The day-ahead price calculations and the balancing (realtime) price
calculations are based on the concept of Locational Marginal Pricing
Separate accounting settlements are performed for each market, the day ahead
market settlement is based on scheduled hourly quantities and on day-ahead
hourly prices, The day-ahead scheduling process will incorporate PJM reliability
requirements and reserve obligations into the analysis. The resulting Day-ahead
hourly schedules and Day-ahead LMPs represent binding financial commitments
to the Market Participants.
The balancing settlement is based on actual hourly (integrated) quantity
deviations from day-ahead scheduled quantities and on real-time prices integrated
over the hour. Generators that are designated PJM capacity resources that are
available but not selected in the day-ahead scheduling may alter their bids for use
in the Real-time Energy Market during the Generation Rebidding Period from
4:00 PM to 6:00 PM (otherwise the original bids remain in effect for the balancing
market). Real-time LMPs are calculated based on actual system operating
conditions as described by the PJM state estimator. LSEs will pay Real-time LMPs
for any demand that exceeds their day-ahead scheduled quantities (and will receive
revenue for demand deviations below their scheduled quantities). Generators are
paid Real-time LMPs for any generation that exceeds their day-ahead scheduled
quantities (and will pay for generation deviations below their scheduled
quantities). Transmission customers pay congestion charges based on Real-time
LMPs for bilateral transaction quantity deviations from day-ahead schedules.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 138

 Figure 6 PJM Scheduling Timeline

The day-ahead scheduling/bidding timeline for two-settlement consists of the

following time frames:
1. 1200 — Day-ahead market bid period closes. All bids must be submitted to
PJM. At 1200 PJM begins to run the two-settlement software to determine
the hourly commitment schedules and the LMPs for the day-ahead market.
This is the first unit commitment run, which determines the unit
commitment profile that satisfies the fixed demand, cleared price-sensitive
demand bids, and PJM operating reserve objectives, while minimizing the
total production cost (subject to certain limitations).
2. 1600 — PJM posts the day-ahead hourly schedules and LMPs on a web-
based Market User Interface (MUI) for the two-settlement system

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 139

 3. 1600 - 1800 — PJM opens the balancing market offer period. During this
time, Market Participants can submit revised offers for units not selected in
the first commitment.
4. 1800 — The balancing market offer period closes. PJM performs a second
unit commitment, which includes the updated offers, updated unit
availability information, and updated PJM load forecast information and
load forecast deviation. The focus of this commitment is reliability and the
objective is to minimize only start-up and no load costs for any additional
resources that are committed. (This analysis could result in the release of
resources that were committed in the day-ahead market).

PJM Regulation Market

The PJM Regulation Market provides PJM participants with a market-based

system for purchase and sale of the Regulation ancillary service.
PJM uses the forecasted LMPs and generation schedules from the Unit
Dispatch System (UDS) to estimate the opportunity cost that would be incurred
by each available unit if it were to provide Regulation during each hour of the
operating day. This estimated opportunity cost is then added to the Regulation
offer to determine a merit order price for each unit for each hour. All available
regulating units are then ranked in ascending order of their merit order prices, and
the lowest cost set of units necessary to meet the PJM Regulation Requirement in
each hour is determined. The highest merit order price associated with this lowest
cost set of units becomes the RMCP for that hour of the operating day
PJM simultaneously optimizes energy, Regulation and Spinning Reserve,
and assigns both Regulation and Spinning to the most cost-effective set of units
each hour of the operating day. If a unit is called on by PJM for the purpose of
providing regulation, the unit is guaranteed recovery of start-up and no-load
costs.

PJM Spinning Reserve Market

PJM uses generation schedules from the Unit Dispatch system to

estimate the amount of incidental Spinning Reserve present on the PJM system
due to economic dispatch, and this capability is designated as Tier 1. Tier 1 is
provided by any unit that is on line, following economic dispatch, and capable of
increasing its output within ten (10) minutes following a call for Spinning Reserve.
If the amount of Tier 1 estimated for a given hour is insufficient to meet the PJM

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 140

Spinning Reserve Requirement, PJM must assign units to operate at a point that
deviates from economic dispatch in order to provide the remainder of the
requirement. The extra capacity that must be committed is designated Tier 2. PJM
uses forecasted LMPs and generation schedules from the Unit Dispatch system
to estimate the opportunity cost (including energy use) that would be incurred by
each available unit if it were to provide Spinning Reserve during each hour of the
operating day. This estimated opportunity cost is then added to the Spinning
Reserve offer to determine a merit order price for each unit for each hour. All
available Tier 2 spinning units are then ranked in ascending order of their merit
order prices, and the lowest cost set of units necessary to meet the PJM Spinning
Reserve Requirement in each hour is determined. The highest merit order price
associated with this lowest cost set of units becomes the SRMCP for that hour of
the operating day.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 141

 ANNEXURE - 6
Overview of SCUC Reporting Software
While analysing the results of SCUC run of 144 ISGS Units over a period of 9 months, need was felt for
a dashboard for visualizing various graphs, parameters and the inputs and results of SCUC run in
addition to having a proper database to archive the data and results. Hence, a software for SCUC
visualization has been developed.

The software is developed with Dash Open Source for visualization, Django Framework as backend,
Postgres ad Database, SCUC using GAMS for SCUC optimization and Redis Open Source as Cache for
faster response of Data visualizations.

SCUC Dashboard:

1
2

Home Page is shown above (labeled with 3). It summarizes various significant parameters of SCUC Run
for the selected Date range. Parameters like SCUC Cost, Startup cost, Savings (Lacs INR) compared to
original schedules and also displays important operational data of Reserves (spinning and cold), SMP
and average costs. Top five toggling units determined by SCUC and region wise summary is also shown.

SCUC Visualization is segregated into various buttons like Summary, SCUC vs Actual, SMP Prices, Unit
status, Reserves, Custom Reserve analysis etc. (labeled with 1).

Common Date Range selector (labeled with 2) is provided across all Tabs for User flexibility.

Summary Page is segregated Plant wise and Unit wise. It includes various Unit Parameters like Unit
Static Data (IC, Region, and Variable cost), Unit Availability Factor (UAF), Unit Capacity Factor (UCF),
Energy in MU and its cost, Startup Cost, No of Hot, Cold and Warm starts for the selected Date Range.
Hovering on any Parameter will provide more information about it (labeled 4). References for various
calculations are placed in Reference Tab.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 142

 4

SCUC vs Actual provides a comprehensive comparison between SCUC Results and Actual Generation.
Comparison can be done individual generator wise, Region wise and All India basis for Schedules,
Spinning Reserves, Cold Reserves, On-bar capacity and Ramping reserves.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 143

SMP Prices compares various price indicators like System Marginal Price (SMP), Average Energy Cost,
and Average cost of reserve etc. SCUC Savings, Heat map of ALL INDIA SMP and Region wise SMP can
also be visualized in this Tab.

Unit Status: Status (RSD, On bar, Outage) of all units is colour coded and displayed for SCUC Run and
for Actual Status.

Reserves: Status of Reserve available for the selected period can be analysed. Reserve Shortfall and
Plant, Unit wise reserves can also be visualized in this Tab.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 144

Custom Reserves Tab helps user to directly interact with SCUC Engine. This allows user to enter
custom reserve requirement block wise. Upon submitting the request SCUC core engine will run in
background and calculate the SCUC Result with specified reserve requirement.

Quadrants: Unit wise classification for any selected day can be viewed using this Tree Map. Units can
be classified into 6 groups (1. Initially On and Committed by SCUC, 2. Initially On and De-committed
by SCUC, 3. Initially Under RSD and Committed by SCUC, 4. Initially Under RSD and De-committed by
SCUC, 5. Toggling Unit, 6. Outage.). Size of each group depends on number of units in that group.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 145

Learning Centre has various references and documents related to SCUC. Users interactive comment
section also provided in which various users can post their queries and suggestion and can be replied
by other users.

ISGS Stack Plot Break up of Generation into Schedule, Reserves and Outage can be visualized in this
Tab.

Raw Data Input and SCUC Results for entire period can be downloaded in zip file.

Miscellaneous Various other miscellaneous plots can be visualized in this Tab like Demand vs SMP, VC
vs ULF, VC vs UCF, Spinning reserves vs SMP, Average Energy vs SMP and Spinning Reserves vs Average
Energy cost.

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 146

 ANNEXURE - 7

Analysis of duals – percentage of time constraints binding in SCED

Apr-Dec 2019 Jan-Dec 2020

S.No. Generators Ramp Ramp Ramp Ramp

Avg VC Pmax Pmin Avg VC Pmax Pmin
Up Down Up Down
(p/kWh) (%) (%) (p/kWh) (%) (%)
(%) (%) (%) (%)
1 Darlipali 109 0 0 99 1
2 Sasan 132 1 0 96 2 135 0 0 98 1
3 Korba-Stage-3 131 0 0 97 2 137 0 0 97 2
4 Rihand 3 133 1 0 96 2 140 0 0 98 1
Korba STPS
5 133 1 1 95 2 140 1 1 95 2
Stage 1 and 2
6 Singrauli TPS 139 2 1 93 3 141 1 1 96 2
7 Rihand 1 134 1 0 96 2 142 1 2 95 2
8 Rihand 2 134 1 0 95 2 142 0 0 96 1
Sipat STPS
9 138 2 2 93 3 145 1 1 95 2
Stage 1
Sipat STPS
10 142 2 2 93 3 149 1 2 94 3
Stage 2
Vindhyachal
11 160 2 3 90 4 160 2 2 91 4
STPS Stage 4
Vindhyachal
12 162 3 4 88 4 162 2 2 89 5
STPS Stage 3
Vindhyachal
13 163 3 4 88 5 164 2 2 88 6
STPS Stage 2
Vindhyachal
14 164 3 4 87 5 165 2 3 88 6
STPS Stage 5
Vindhyachal
15 170 4 5 84 6 170 3 4 84 7
STPS Stage 1
16 CGPL Mundra 189 4 5 80 9 195 1 0 84 15
Talcher Stage
17 200 3 4 79 11 203 3 3 77 12
2-NTPC
18 Talcher 196 6 7 76 12 204 4 4 76 13
19 NPGC 219 8 10 68 13 208 4 5 79 12
Kahalgaon
20 213 5 6 74 12 215 3 5 75 17
Stage 2
21 LARA 237 7 7 67 19 224 4 8 72 16
Kahalgaon
22 224 4 5 69 18 225 4 4 66 22
Stage 1
23 Nabinagar 212 6 7 72 14 239 5 5 59 27
24 Tanda-2 250 4 5 57 34
Ramagundam
25 260 8 8 53 28 251 4 5 74 17
TPS Stage 3
Ramagundam
26 264 7 7 43 33 254 4 5 50 29
TPS Stage 1&2
Neyveli TPS 1
27 244 5 4 62 28 254 0 0 61 38
(Expn.)-NLC
Neyveli TPS 2
28 249 2 4 69 25 259 0 1 61 38
(Expn.)-NLC
Farakka Stage
29 241 7 8 58 21 266 5 5 29 53
1 and 2

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 147

 Apr-Dec 2019 Jan-Dec 2020

S.No. Generators Ramp Ramp Ramp Ramp

Avg VC Pmax Pmin Avg VC Pmax Pmin
Up Down Up Down
(p/kWh) (%) (%) (p/kWh) (%) (%)
(%) (%) (%) (%)
FARAKKA-
30 239 4 4 73 19 266 3 3 42 53
STAGE-3
31 MPL 270 14 16 32 36 266 7 7 36 46
32 MTPS Stg 2 273 10 11 37 40 271 6 7 31 53
33 Jhabua 271 2 2 16 34
34 Gadarwara-1 349 6 4 54 36 274 5 7 67 20
35 Barh 231 8 9 61 19 277 6 6 28 54
Neyveli-I
36 266 5 4 49 38 277 0 0 35 63
(ISGS)-NLC
Neyveli-II
37 266 6 4 49 38 277 0 1 41 58
(ISGS)-NLC
NTPC-
38 278 6 6 44 44
KHARGONE
39 NSPCL 334 11 11 20 57 283 8 7 32 50
40 NTPL 314 11 10 21 55 284 7 7 23 60
41 Mauda 313 12 11 25 50 295 7 9 29 54
42 Mauda 2 315 13 12 28 47 297 7 13 29 50
43 NTPC SOLAPUR 368 6 6 38 50 302 4 5 46 45
Simhadri NTPC
44 322 9 8 23 55 306 4 15 23 58
Stage 2
Simhadri -NTPC
45 325 8 9 18 59 310 5 5 11 76
Stage -1
46 UNCHAHAR 4 311 11 12 30 47 311 5 8 14 72
47 Unchahar TPS 1 331 9 9 19 60 330 5 4 9 81
48 Unchahar TPS 3 331 9 9 18 60 330 4 4 7 83
49 Unchahar TPS 2 333 10 9 16 63 333 5 4 6 84
50 Kudgi-1 388 5 5 9 81 336 2 29 33 35
51 Bongaigaon 325 11 11 17 57 344 4 4 4 86
Indira Gandhi
52 364 7 6 12 73 346 4 8 11 77
TPS
53 Vallur NTECL 381 6 5 3 82 347 3 8 10 79
54 Dadri Stage 2 377 6 5 6 78 355 2 10 10 78
55 Dadri Stage 1 405 4 3 1 88 372 1 34 35 31

Note: The reported percentages have been normalized with respect to the share of time when station
was on-bar

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 148

 ANNEXURE - 8
Computation of Location Marginal Prices (LMP) for
Indian Grid
Motivation

The All India grid is a well interconnected network. However, there are zones and flow
gates where delivery of power across the seam could pose challenges in certain
conditions. SCED & SCUC do not consider network implicitly. ATC of flow gates/Bid
areas is considered as a pre decided constraint.

To take care of congestion in areas/zones which are not at point of the flow gates
where ATC/TTC is declared beforehand and also to provide a techno-economic
indication, LMP would be a good indication.

The LMP at every node would reflect the cost of delivering power to that node
considering the system marginal price, loss factor and the congestion price. This could
be an input to the SCUC & SCED for further refinement.

Locational Marginal Pricing

Locational Marginal Pricing (LMP) is the cost to supply the next increment of demand
at each node in the system while respecting the binding transmission constraints. The
LMPs are be used to (a) Establish price for energy purchases and sales at specific
locations, (b) To pay generators/suppliers for the power supplied, (c) To collect charges
from consumers/load for the power consumed and (4) for collecting transmission
congestion charges. The LMP at any bus ‘k’ consists of three (3) components

i. System Marginal Price (SMP)

ii. Loss component which depends on the loss sensitivity to drawal at bus k
iii. Congestion component which depends on the line flow sensitivity to drawal at
bus k

Sample Case Study of Indian grid

LMPs for each node/substation in the Indian grid are determined using Optimal Power
Flow (OPF). OPF is an optimization problem consisting of an objective function with
equality and inequality constraints. The objective function may consist of fuel cost
optimization, loss optimization etc. The equality constraints include Power flow
equations at each node and the inequality constraints include Pmax/Pmin of
generators, bus voltages magnitude and angles etc. LMPs at each node in the All India
grid are determined with the following inputs to OPF:

(i) All India Basecase

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 149

 (ii) Variable cost of each generator
(iii) Transmission constraints such as line loading limits, Voltage limits

The results of the OPF study with LMPs at different nodes without and with congestion
towards Kerala are shown in Figure A. Without congestion in the grid, the prices at
different nodes are almost same. The small difference in the price at different prices is
due to the losses in the system and marginal generator cost generator is Mettur TPS
(unit cost Rs. 3.49/-) which is located in Tamilnadu. With congestion towards Kerala,
the prices in the Kerala has gone up from Rs. 3.55 to Rs. 8.00 and the costly generator
in Kerala i.e., Kayankulam (unit cost Rs. 6.00/-) also gets dispatched.

Figure A LMPs at different nodes without and with congestion

SCED Pilot - Detailed Feedback Report – Mar 2021 Page 150

Power System Operation Corporation Limited
(A Government of India Enterprise) | CIN: U40105DL2009GOI188682

B-9, Qutab Institutional Area, New Delhi – 110016

www.posoco.in | posococc@posoco.in