Research Proposal Submitted to Grid-India from IIT Kharagpur

1. Project Title: Development of a support tool for renewable energy resources redis-
patch from load despatch centres.
2. Project Lead: Dr. Deepak Reddy Pullaguram (Assistant Professor, Department of
Electrical Engineering, IIT Kharagpur)

3. Co-Ordinator(s) of Project:
Dr. Ashish Ranjan Hota (Assistant Professor, Department of Electrical Engineering,
IIT Kharagpur)
Prof. Ashok Kumar Pradhan (Professor, Department of Electrical Engineering, IIT
Kharagpur)

4. Motivation:
In modern power systems, increasing integration of renewable energy sources (RES)
such as wind and solar power has led to reduced system inertia. As a result, sys-
tem frequency deviations and maintaining stability have become a growing concern,
especially in grids dominated by renewable energy, where variability in power out-
put can lead to rapid frequency fluctuations. Extreme weather conditions, such as
heatwaves, storms, or prolonged periods of low wind or sunlight, exacerbate these
challenges. When renewable generation exceeds demand, the grid may experience
over-frequency conditions; conversely, when renewable output drops unexpectedly,
under-frequency issues may arise. Both scenarios, if not promptly addressed, can
compromise grid reliability and potentially lead to widespread outages. To sum-
marize, growing share of renewable generation have introduced the following chal-
lenges for grid operators.
• Reduction in Inertia: Conventional generators provide rotational inertia that
naturally resists sudden frequency drops. The net inertia will be reduced with
increased RES penetration, and can lead to frequent fluctuations in frequency.
• Rapid Frequency Deviations: The lack of immediate response from RES re-
sults in a higher rate of change of frequency (RoCoF), which necessitates fast
frequency response mechanisms.
• Stochastic Nature of Renewables: Wind and solar power exhibit variability and
uncertainty, requiring stochastic optimization-based dispatch to ensure system
reliability.
• Congestion Management: High RES integration and their intermittency can
lead to transmission congestion, making redispatch essential to maintain secu-
rity constraints.
To mitigate the above risks, transmission system operators (TSOs) employ redis-
patch—a dynamic process that adjusts power plant outputs in near real-time to pre-
vent congestion and balance the grid. After receiving the day-ahead generation

1
 schedules, TSOs perform detailed load-flow and frequency analyses, taking into ac-
count factors such as unplanned outages, weather-induced variability, and mainte-
nance activities. If the analysis indicates that parts of the grid will likely experience
over- or under-frequency conditions, the TSO instructs specific power plants to mod-
ify their output. This may involve curtailing generation in areas where excess power
is driving up the frequency or ramping up production in regions facing a deficit,
thereby restoring balance without altering the total energy supplied.
Given the increasing penetration of renewable energy and the complex interplay of
weather-related impacts, there is a growing need for an automated, real-time tool to
perform comprehensive redispatch analysis. Such a tool would continuously mon-
itor grid conditions, predict potential frequency and congestion issues, and quickly
communicate redispatch instructions to all relevant parties—including renewable
generation plants, fast-acting storage systems, and demand response providers. Au-
tomation would enable rapid response even for transient events and extreme weather
scenarios, ensuring adjustments occur within the critical time frames necessary to
stabilize the system. This project aims to develop such a tool as detailed in the fol-
lowing sections.
5. Objective: The objective of this research proposal is to address the above-mentioned
challenges as described below.
(a) Develop a Robust Redispatch Model: Formulate an optimal power flow-based
redispatch model that considers both frequency and line security constraints
while ensuring economic dispatch constraints.
(b) Ensure Frequency Stability: Integrate frequency security constraints, such as
RoCoF and frequency nadir limits, to prevent large frequency excursions that
could lead to system instability.
(c) Real-time dynamic redispatch: Develop dynamic congestion and frequency man-
agement technique using dynamics model and sensitivity matrices that used for
redispatch actions and avoid overload transmission lines during any schedule
deviations in the system.
(d) Enhance Computational Efficiency: Develop scalable solution methodologies,
to solve the redispatch problem in real time.
A brief overview of the related literature is presented in the following section, fol-
lowed by description of research gaps and detailed methodology.
6. Review of Existing Works and Research Gap:
Redispatch is widely used in the literature to provide for optimal generation alloca-
tion [1], to fulfill spinning reserve requirements [2], to meet stability requirements [3],
to reduce generator rotor shaft impacts induced by transmission loop closures [4],
and to improve resiliency [5], among other objectives. The redispatch problem in-
volves optimally allocating system generation among generating units while track-
ing a load curve and observing the power rate limits of the units, spinning reserve
requirements, and other security constraints. For example, in [3] the impact of gen-
eration redispatch variations on system stability margins with respect to Hopf bi-
furcations and robustness against uncertainties in load is studied. A novel genera-
tion rescheduling method to reduce generator rotor shaft impacts from transmission

2
loop closures is presented in [4]. In [2], a quadratic programming technique com-
bined with a linear programming redispatch method—as well as a Dantzig-Wolfe
decomposition-based approach is proposed. An analytic formula for modal sensi-
tivity is developed in [6] to rank the best pairs of generators for redispatch in order
to damp inter-area oscillations. More recently, synchrophasor measurement-based
real-time congestion management using generator and battery energy storage sys-
tem (BESS) dispatch has been proposed in [7]. In that approach, a dynamic model of
line power flow is used and Model Predictive Control (MPC)-based compensation is
developed to regulate the line power flow at or below the desired value by account-
ing for output and rate limits and maintaining active power balance in the network.
Additionally, a real-time redispatch strategy for energy storage systems—designed
to enhance system reliability by mitigating loss-of-load events while largely adher-
ing to market schedules is proposed in [8], formulated through logical conditions and
integrated into a Monte Carlo-based stochastic adequacy model for multiple battery
configurations in the Greek power system.
However, most of these redispatch techniques are developed to satisfy specific re-
quirements, such as stability, resiliency, congestion, or contingency management.
A critical analysis of existing approaches to balancing and congestion management
(specifically redispatch) and their effects on service provider incentives is needed [9].
Redispatch becomes crucial in grids with high renewable penetration due to the in-
termittency of these sources; renewable energy sources are often curtailed when con-
ventional options are exhausted or when grid security requirements dictate.
In Germany, redispatch is regulated by the Energy Industry Act (EnWG) as of Octo-
ber 1, 2021. Under Redispatch 2.0, this regulation has expanded to include all gen-
eration units of 100 kW or more—including renewable energy, energy storage, and
CHP plants—with the aim of improving grid integration, enhancing efficiency, and
reducing redispatch costs while ensuring non-discriminatory congestion manage-
ment [10, 11]. Under this regime, renewable energy sources are curtailed only when
conventional options are exhausted or when significant cost savings can be achieved;
compensation follows a cost-based approach per Section 13a of EnWG, ensuring op-
erators remain financially neutral, although participation is limited due to the lack
of profit incentives and redispatch costs are ultimately passed on to electricity con-
sumers via grid charges. Similarly, in California, redispatch is used as a real-time
tool by the CAISO to manage transmission congestion; when scheduled flows ex-
ceed physical limits, the CAISO adjusts generation output by calling on generators
with the lowest decremental bids to ramp up or down to maintain grid balance.
As India expands its renewable integration, grid security is increasingly at risk, es-
pecially during high renewable seasons when significant frequency deviations occur
over extended periods. Frequency and line congestion management based on redis-
patch are therefore essential for stabilizing grid frequency and preventing transmis-
sion overloads. This project aims to develop two redispatch strategies: (i) frequency-
constrained optimal power flow-based generation redispatch and (ii) real-time dy-
namic line congestion and frequency management by developing the line and fre-
quency sensitivity matrix with respect to generation, energy storage systems, and
renewable generation.

3
7. Methodology: The project objectives stated earlier will be pursued in two work
packages (WPs) as detailed below.
• WP1: Frequency-Constrained Optimal Power Flow
The redispatch problem aims to minimize the cost of generation adjustments
while ensuring frequency stability, and to avoid transmission line congestion.
Specifically, the goal is to determine the deviations for each generator from their
scheduled dispatch to meet the power balance and minimize frequency devia-
tion. In order to compute an economic redispatch, the following optimization
problem is formulated:
X up X up
min CGi ∆PGupi + CGdown
i
∆P down
Gi
+ CDi ∆PDupi + CD
down
i
∆PDdown
i
(1)
∆P up ,∆P down
i∈G i∈BDR

where, CGupi , CGdown

i
are the Cost coefficients for increasing ∆PGupi or decreasing
up
∆PGdown
i
power output of generator i ∈ G during redispatch. CD i
down
, CD i
are the
up down
Cost coefficients for increasing ∆PDi or decreasing ∆PDi power output of en-
ergy storage system or load with demand response at bus i ∈ BDR during re-
dispatch. Here, G and BDR are the set of generators and buses with demand
response, respectively.
To ensure the frequency deviation is within the permissible limits during gener-
ation and load deviations from scheduled values, the following additional con-
straints are imposed on the optimization problem.
Rate of Change of Frequency (RoCoF) Constraint: The maximum permissible RoCoF
ensures frequency stability in response to disturbances:
∆Pk
P P ≤ RoCoFmax ∀k ∈ B, (2)
i∈G Hi Si + j∈R Mv,j Sj

where: Hi , Mv,j : Inertia constants of conventional and renewable generators,

Si , Sj : Apparent power ratings, ∆Pi change in generation/load from the sched-
uled, R, is the set of Renewable generator.

Frequency Nadir Constraint: The frequency deviation must remain within the ac-
ceptable bounds:
 
∆Pk M Rg ∆Pk
∆fnadir = − + ln 1 + ≥ fnadir, min , (3)
D D2 M Rg
where:
– D: Load damping coefficient.
– M : System inertia.
– Rg : total droop coefficient for based Available governor response.
This work package formulates the fundamental frequency-constrained optimal
power flow model, incorporating the system droop constant D, which defines
the relationship between net power changes and frequency deviations and sys-
tem inertia H for obtaining the rate of change of frequency. Considering the

4
 power system having a set of buses (B), generators (G), and lines (L), then the
frequency-constrained optimal power flow is formulated as

X

min Ci ∆PGi (4a)
i∈G
X  
s.t. PGnew
i
− PDnew
i
= Vi Vj Gij cos(θi − θj ) + Bij sin(θi − θj ) , ∀i ∈ B
j∈N (i)
(4b)
X  
Qnew new
Gi − QDi = Vi Vj Gij sin(θi − θj ) − Bij cos(θi − θj ) , ∀i ∈ B
j∈N (i)
(4c)
PGmin
i
≤ PGnew
i
≤ PGmax
i
, ∀i ∈ G (4d)
new
Qmin max
Gi ≤ QGi ≤ QGi , ∀i ∈ G (4e)
Vimin ≤ Vi ≤ Vimax , ∀i ∈ B (4f)
Pij2 + Q2ij ≤ (Sijmax )2 , ∀(i, j) ∈ L (4g)
!
1 X X
∆PGi − ∆PDi ≤ RoCoFmax (4h)
2H i∈G i∈B
P P

i∈G ∆PGi − i∈B ∆PDi
∆fmin ≤ − 1 ≤ ∆fmax (4i)
D+ β

(4j)
where ∆PGi and ∆PDi are the change in the Generation and Load with demand
response for the dispatch. Here, (4g) is the line congestion constraint, and (4h)
and (4i) are the additional frequency constraints. In this, PGnew
i
, Qnew
Gi are the gen-
eration variables for i ∈ G, PDnew
j
Qnew
Dj are the demand variables associated to the
loads that participate in demand response i ∈ BDR , Vi are the voltage variables
corresponding to all buses B. The frequency-constrained OPF is solved when-
ever the frequency deviates for more than two consecutive block periods. These
new variables are sent to all generating and renewable units, loads with demand
response.
• WP2: Dynamic Power flow line congestion management
This work package aims to develop dynamic real-time line congestion and fre-
quency management by generation re-dispatch. Although the frequency-constrained
OPF provides the generation and load set-points, these setpoints are updated
for every 15-minute block period. However, real-time dispatch, in conjunc-
tion with the OPF, will be developed, where any expected deviations in gen-
eration/demand are handled. In this work, a detailed state-space model that
captures generator and Battery Energy Storage System (BESS) dynamics will be
developed, and the output matrix corresponding associated with line-flows and
frequency will be obtained.

5
Continuous-Time State-Space Model: Defining the overall state vector as
 
xg (t)
xc (t) = ,
xb (t)

where:
– xg (t) ∈ Rng represents the generator states (including deviations in frequency
∆ωi (t), rotor angle ∆δi (t), valve position ∆P SVi (t), turbine power ∆P CHi (t),
and mechanical power ∆P Mi (t) for each generator i).
– xb (t) ∈ Rnb represents the BESS states (e.g. deviations in battery power injec-
tion ∆P Bi (t) for each unit).
The continuous dynamics are modeled by:
ẋc (t) = Ac xc (t) + Bc uc (t) + Bc1 uc1 (t) + dc (t), (5)
with the output equation

yc (t) = Cc xc (t) + Dc1 uc1 (t). (6)

Here:
– uc (t) is the control input vector (e.g., generator re-dispatch commands ∆PCi (t)
and BESS set-points ∆PB∗ i (t)).
– uc1 (t) represents external disturbances (such as fluctuations from renewable
sources).
– dc (t) aggregates modeling uncertainties and other disturbances.
– Cc is the line flow sensitivity matrix with respect to generation, BESS, and
buses with Demand response injection, and Dc1 is the line flow sensitivity
matrix with respect to renewable injection

∂Pi,j
elements of Cc = ; ∀{i, j} ∈ L, ∀k ∈ G ∪ BDR (7)
∂PPk
∂Pi,j
elements of Dc1 = ; ∀{i, j} ∈ L, ∀k ∈ R (8)
∂PPk

Assuming the transmission line power flow is represented by yd and its safe limit
as r, the frequency as ω and ωref as reference value, then we define the tracking
error as,  
yd (t) − r
e(t) = (9)
ω(t) − ωref
An augmented state is then formed,
 
∆xc (t)
x(t) = ,
e(t)

leading to the augmented model,

x(t) = A x(t) + B u(t) + D dd (t), (10)

6
 Year1 Year2
0 − 6 Months 6 − 12 Months 12 − 18 Months 18 − 24 Months
State-of-the-art
WP1: Frequency OPF Problem
WP1: Optimal Redispatch solving tool
WP2: Dynamic re-dispatch problem
WP2: computationally efficient solution
Deployment and corrective actions

Figure 1: Work flow Timeline

with output
y(t) = C x(t).
Using, the PMU measurments, by applying Kalman filter for state estimation,
an optimal state feedback control will be developed to optimally redispatch the
generation and renewable curtailment in real time.
8. Work plan and Time Schedule
(a) Organization of Work Elements: The PIs will recruit a junior research fellow
(JRF) to work on this project. In order to start making early progress in the
project, the PIs will also involve one existing PhD scholar on full time basis.
The honorarium available via the project will enable the PIs to attract and retain
students of exceptional caliber in the project. In addition, the PIs will provide
regular updates to Grid-India about their progress, and incorporate their feed-
back in different work packages.
(b) Time Schedule of Activities: A Gantt chart of the sequence of major activi-
ties/milestones is given in Figure 1.
9. Expected Outcome/Deliverable: The following outcomes are expected.

• This project is expected to study the existing redispatch approaches adopted

across the globe.
• A robust framework for a frequency-constrained optimal power flow-based re-
dispatch strategy incorporating detailed all security constraints and other con-
straints associated with generators, energy storage systems, and renewable sources.
• A real-time dynamic congestion and frequency management algorithm that uti-
lizes sensitivity matrices to adjust generation and storage dispatch.
• The developed tool will be deployed and corrective actions will be taken follow-
ing active feedback from Grid-India.

7
10. Proposed Budget

Item Budget Justification Amount

(in INR)
A Recurring
1. Salaries/wages Honorarium/top-up salary @ Rs. 20,000 14,56,800
per month to one existing PhD student,
salary @ Rs. 37,000 per month to one JRF
2. Consumables Purchase of books, Journal over-length 9,00,000
charges, Up-gradation of lab. computing
facilities, Proprietary optimization solvers
(such as Mosek or Gurobi)
3. Travel National and international conference 3,00,000
travel, travel for data collection, project
meetings travel
4. Other costs/ Con- Conference registration fee, project meet- 2,43,200
tingencies ing expenses.
B Equipment One Laptop and one workstation 4,00,000
C Overhead 20% of (A+B) as per IIT norms 6,60,000
Grand Total (A+B+C) 39,60,000

Break-up of Budget for Salaries/Wages:

Designation and number Monthly Budget for Budget for Total

of persons Salary Year 1 Year 2 Amount
(in INR)
Top up salary/honorarium 20,000 2,40,000 2,40,000 4,80,000
for one PhD student
Salary for one JRF, includ- 40,700 4,88,400 4,88,400 9,76,800
ing HRA
Total 60,700 7,28,400 7,28,400 14,56,800

List of equipment available with the PI’s institute to be used under this project:
The institute will provide lab facilities including two Desktop computers, Real-time-
simulator, printers, scanners, tablets, monitors that facilitate online collaborations,
among others.
11. List of relevant publications by the investigators
(a) Venkateswara Rao Kagita, Sanjaya Kumar Panda, Ram Krishan, P. Deepak Reddy
and Jabba Aswanth, “High-performance computing for static security assess-
ment of large power systems," Connection Science, vol. 35, no.1, 2023.
(b) Deepak Pullaguram, R. Madani, T. Altun and A. Davoudi,“Optimal Power Flow
in AC/DC Microgrids with Enhanced Interlinking Converter Modeling,” in IEEE

8
 Journal of Emerging and Selected Topics in Industrial Electronics, vol. 3, no. 3, pp.
527-537, July 2022.
(c) Deepak Pullaguram, Ramtin Madani, Tuncay Altun, and Ali Davoudi, “Small
Signal Stability Constrained Optimal Power Flow for Inverter Dominant Au-
tonomous Microgrid", in IEEE Transactions on Industrial Electronics, vol. 69, no. 7,
pp. 7318-7328, July 2022.
(d) A. Mishra, A. Mallick, A. R. Hota, D. R. Pullaguram and P. Bajpai, “A Trans-
active Energy Framework for Real-Time Primary Frequency Support by Active
Distribution Networks", 2024 IEEE PES Innovative Smart Grid Technologies -
Asia, Bangalore, 2024.
(e) B. K. Poolla, A. R. Hota, S. Bolognani, D. Callaway and A. Cherukuri, “Wasser-
stein Distributionally Robust Look-Ahead Economic Dispatch," IEEE Transac-
tions on Power Systems, vol. 36(3), pages: 2010 - 2022, 2021.
In particular, PI Hota and PI Pullaguram have collaborated in the past, and have a
joint publication on this topic.
12. Details of Research Projects being implemented/ completed/ submitted by the in-
vestigator Dr. Deepak Reddy Pullaguram
(a) “Development and Testing of Cyber Resilient strategies for Converter Dominant
Microgrid Systems," ISIRD, SRIC, IIT Kharagpur, Sept 2023 - Sept 2026 (INR
25,00,000).
(b) “High sensitive grounded conductor identification for human safety in low ten-
sion (LT) system," CPRI Banglore, July 2023- Oct 2024, (INR 21,56,000) (Co-PI)
Completed.
(c) “Convex optimization based economic and stable operation of microgrid sys-
tem,” Start-up Research Grant (SRG) Sponsored by SERB-DST, Dec 2020 - Dec
2022, (INR 32,04,860) Completed.
(d) “Dynamic and Static Security Analysis of Large Power System using High- Per-
formance Computing,” HPC Applications, Sponsored by National Supercomput-
ing Mission (NSM)-DST, Mar 2021 - Mar 2023, (INR 45,58,160) (Co-PI) Com-
pleted.
(e) “Stability Improvement of Microgrid System with Uncertain Generation and
Loads,” Research Seed (RSM) sponsored by National Institute of Technology
Warangal June 2020 - June 2022, (INR 19,99,500) Completed.

References
[1] R.-H. Liang, “A neural-based redispatch approach to dynamic generation allocation,” IEEE Transac-
tions on Power Systems, vol. 14, no. 4, pp. 1388–1393, 1999.
[2] C. Somuah and N. Khunaizi, “Application of linear programming redispatch technique to dynamic
generation allocation,” IEEE Transactions on Power Systems, vol. 5, no. 1, pp. 20–26, 1990.
[3] A. Zecevic and D. Miljkovic, “The effects of generation redispatch on hopf bifurcations in electric
power systems,” IEEE Transactions on Circuits and Systems I: Fundamental Theory and Applications, vol. 49,
no. 8, pp. 1180–1186, 2002.

9
 [4] N. Martins, E. J. de Oliveira, W. C. Moreira, J. L. R. Pereira, and R. M. Fontoura, “Redispatch to reduce
rotor shaft impacts upon transmission loop closure,” IEEE Transactions on Power Systems, vol. 23, no. 2,
pp. 592–600, 2008.
[5] M. Abdelmalak and M. Benidris, “Proactive generation redispatch to enhance power system resilience
during hurricanes considering unavailability of renewable energy sources,” IEEE Transactions on In-
dustry Applications, vol. 58, no. 3, pp. 3044–3053, 2022.
[6] S. Mendoza-Armenta and I. Dobson, “Applying a formula for generator redispatch to damp interarea
oscillations using synchrophasors,” IEEE Transactions on Power Systems, vol. 31, no. 4, pp. 3119–3128,
2016.

[7] K. Chakravarthi, P. Bhui, N. K. Sharma, and B. C. Pal, “Real time congestion management using gener-
ation re-dispatch: Modeling and controller design,” IEEE Transactions on Power Systems, vol. 38, no. 3,
pp. 2189–2203, 2023.
[8] P. A. Dratsas, G. N. Psarros, and S. A. Papathanassiou, “A real-time redispatch method to evaluate the
contribution of storage to capacity adequacy,” IEEE Transactions on Power Systems, vol. 39, no. 1, pp.
1274–1286, 2024.
[9] K. Poplavskaya, M. Joos, V. Krakowski, K. Knorr, and L. de Vries, “Redispatch and balancing: Same
but different. links, conflicts and solutions,” in 2020 17th International Conference on the European Energy
Market (EEM), 2020, pp. 1–6.
[10] “Redispatch 2.0: Important information for plant operators,” https://www.netztransparenz.de/en/Ancillary-
Services/System-operations/Redispatch/Redispatch-20-at-the-TSO-DSO-interface, 2024, [Online].
[11] F. Billault-Chaumartin, M. Eising, and A. Motte-Cortés, “Redispatch patterns of energy market and
reserve power plants in germany,” in 2020 17th International Conference on the European Energy Market
(EEM), 2020, pp. 1–6.

10