Forecasting and scheduling of wind and solar power generation in india

Session: Reduce/Renewable/Reuse/
Retrofit/Rebuild
GETS 2016 ID NO #
Forecasting and Scheduling of Wind and Solar Power generation in India
Brief Abstract: Wind and solar energy are the major components of renewable energy
in India but the variability and unpredictability inherent to wind and solar power raise a
number of issues associated with grid integration. The integration of significant wind and
solar energy into existing supply system is a challenge for large scale renewable energy
penetration; hence the day-ahead and short-term renewable energy forecasting is needed
to effectively integrate renewable power to the grid. As per CERC regulation, FOR
regulation and different state regulations, day-ahead forecasting and scheduling of wind
and solar energy is required to be submitted by IPPs and hence the forecasting and
scheduling of renewable energy has become a widely pursued areas of research at
Indian context. Though there are different methods in forecasting renewable power
generation, in this paper, we are showing how the mixed approach technique using
artificial intelligence, the algorithm of which are developed by del2infinity, is useful in
forecasting of wind and solar power generation. Since the regulation deals with the
deviation settlement mechanism due to erroneous forecast results, this paper also shows
the theoretical structure and simple penalty calculation methodology using probabilistic
model for different scenarios of forecast accuracy. The approximate statistical model
correlating forecast accuracy and penalty due to erroneous forecast can act as an
essential tool in maximizing the energy accuracy to minimize the penalty due to deviation.
Author: Abhik Kumar Das
Abhik holds a Dual Degree (B.Tech in Electronics & Electrical
Communication Engineering & M.Tech in Automation and
Computer Vision) from Indian Institute of Technology, Kharagpur,
India. He has a vast experience in computational modelling of
complex systems. He contributed in different verticals of analytical
modelling related to renewable energy and techno-economics and
published several well-cited research articles in internationally
acknowledged journals and peer-reviewed conferences. He is founding member of
del2infinity, an accurate Wind Energy & Solar Energy Forecasting and Scheduling
Solutions Company. (e-mail: contact@del2infinity.xyz).

Distributed with permission of author(s) by GETS 2016. Published at GETS 2016 http://www.NTPCGETS.com

Abstract— Due to the non-availability of sufficient resources and considerable amount of emission of pollutants from
commercial energy generation, renewable energy has been high on Indian policymakers’ agenda for enabling the country`s
transition to a sustainable energy future. Wind and solar energy are the major components of renewable energy in India but the
variability and unpredictability inherent to wind and solar power raise a number of issues associated with grid integration. With
increasing penetration of wind and solar power having unscheduled fluctuations, wind and solar power creates a threat to grid
reliability due to balancing challenge in load and generation. The integration of significant wind and solar energy into existing
supply system is a challenge for large scale renewable energy penetration; hence the day-ahead and short-term renewable
energy forecasting is needed to effectively integrate renewable power to the grid. As per CERC regulation, FOR regulation and
different state regulations, day-ahead forecasting and scheduling of wind and solar energy is required to be submitted by IPPs
and hence the forecasting and scheduling of renewable energy has become a widely pursued areas of research at Indian
context. Though there are different methods in forecasting renewable power generation, in this paper, we are showing how the
mixed approach technique using artificial intelligence, the algorithm of which are developed by del2infinity, is useful in
forecasting of wind and solar power generation. Since the regulation deals with the deviation settlement mechanism due to
erroneous forecast results, this paper also shows the theoretical structure and simple penalty calculation methodology using
probabilistic model for different scenarios of forecast accuracy. The approximate statistical model correlating forecast
accuracy and penalty due to erroneous forecast can act as an essential tool in maximizing the energy accuracy to minimize the
penalty due to deviation.
Index Terms— Wind, Solar, Forecasting, Penalty, Energy Accuracy
I. INTRODUCTION
Energy is a vital requirement for social and economic development of a nation. One of the major indices of
improved quality of life is per capita energy consumption which has been rising steadily in India, though a
considerable amount of villages are without power even today [1]. Due to the socio-economic development, the
demand of energy has multiplied manifold and this demand can be no longer satisfied by the traditional energy
technology using local resources only and to envision energizing the rural and urban area with high energy
demand, renewable energy is being seen as a transformative solution to meet energy demand as well as
economic challenges. For a sustainable energy future, not only the energy demand but amount of emission of
pollutants from commercial energy generation is a crucial issue; and considering National Action Plan on Climate
Change [2], Nationally Determined Contributions (NDCs) [3], renewable energy (RE) goals [4], India‟s national
policies and other initiatives encourage renewable and clean energy for various applications. At present, India has
a target of 175 Giga Watt (GW) of installed capacity from renewable energy by 2022, of which 100GW is to come
from solar, 60GW from wind [4]. In addition, India‟s NDC goal is to achieve 40% of total installed power generation
capacity from renewable energy by 2030 [3] and therefore, a great interest in adopting green energy technologies
in the country [5]. However, the large-scale deployment of renewable energy technology involves a combination of
interventions involving policy and regulatory mechanisms, technological solutions and institutional structures.
Abhik Kumar Das is Director of del2infinity Energy Consulting Pvt. Ltd., India. He is a B.Tech & M.Tech (Dual Degree) from IIT Kharagpur
and working in the domain of computational modeling of complex systems for last 12 years. (e-mail: contact@del2infinity.xyz).
Forecasting and Scheduling of Wind and Solar Power generation in India
Abhik Kumar Das

The conversion of wind energy to the electrical energy is generally done by wind turbine. The wind stream
produces aerodynamic forces on the turbine blades to rotate and capture the kinetic energy contained in the wind.
The captured energy depending on power curve [6-8] of the turbine is transferred through a gearbox to an
electrical power generator, which sends the power into the electrical grid system. Similarly, depending on current
voltage characteristics of solar cells [9-11], the solar energy is converted into electrical energy which is transferred
in to the electrical grid system. Since the energy generated using wind and solar is depended on natural
phenomena and shows stochastic behaviour in generation patterns [12, 13], the wind and solar power generated
is intermittent in nature and highly variable depending on the different parameters of nature. Due to the
intermittency and variability, wind and solar energy shows ramping patterns in renewable energy generation and
hence the large penetration of renewable energy creates instability in existing grid unless proper forecasting and
scheduling is performed.
The demand supply characteristics of complex grid network are a crucial issue for power management. Utility
system planning and operations for both generation and transmission can be affected by wind and power with its
increasing penetration. Since wind and solar power are inherently intermittent and comes with large amount of
unscheduled fluctuations, balancing load and generation creates a great threat to grid reliability. Forecasting is an
essential requirement of variable renewable energy for grid stability as the major purpose of forecasting is to
reduce the uncertainty of renewable generation, so that its variability can be more precisely accommodated.
The concept of forecasting and scheduling of renewable energy generators and the commercial settlement was
introduced in Indian context by CERC through Indian Electricity Grid Code (IEGC), 2010 [14] and the Renewable
Regulatory Fund mechanism [15] was envisaged to be implemented from January 1, 2011. Due to several
implementation issues, the mechanism was never made operational. To formulate an implementable framework,
CERC on 31.03.2015 issued a draft Amendments. Based on comments and suggestions received from various
stakeholders, CERC published the third amendment to IEGC which is issued on 07.08.2015. On the same date
CERC also issued 2nd amendment to regulation for Deviation settlement mechanism and other related matters
[16]. Since the system operators in India have to do curtailment on variable renewable energy due to intermittency
and variability of the wind and solar power generation, the forecasting takes an important role in creating a
sustainable solution for maximum utilization of renewable energy. After CERC regulation, Forum of Regulators
(FOR) [17] and other state regulators issued or drafted regulation related to the forecasting and scheduling of
Wind and Solar power. Apart from regulatory framework, this paper concentrates on the forecasting methodology
and penalty due to regulations related to forecasting and scheduling of wind and solar energy.
The remaining of the paper is organized as follows: In Section II, the variability of wind and solar energy is
discussed using empirical functional relationship. Section III describes the accuracy of forecasting and scheduling
and its effect on penalty due to deviation considering the regulation. The forecasting methodology and brief results
are shown in Section IV and the conclusion is presented in Section V.
II. VARIABILITY OF WIND & SOLAR ENERGY
Wind and solar energy is stochastic in nature and while discussing about the planning and operation of the power
grid, variability and uncertainty plays a crucial role. Variability of power represents the change of generation output
due to fluctuations of wind or sun while uncertainty describes the inability to predict in advance the changes in

generation output. Large unscheduled changes in wind or solar output are called ramp events and they hamper
the penetration of variable power in the existing grid. The variability can be quantified as a measure of dispersion
in variable renewable power generation and be easily quantified using the concept of Lorenz curve [18]. Though
there exists different ways to quantify variability, the ratio based model can be useful to measure variability for
computational feasibility. The analytical representation of variability depends on the stochastic variable
defined as [12],
(1)
for t = 0, 1, 2,….. T – Δt. Here P(t + Δt) is the power generation at a time Δt ahead of t and Δt >0. If P(t) = 0, we
can assume that the value is defined and can take a very high value ( tends to infinity ) and the minimum
value of is 0 as . Since the ramp up (ramp-down) events are defined when the rate power
change is positive (negative) [19-23], using we can say that,
for ramp-down events (2.A)
for ramp-up events (2.B)
Let consider the cumulative distribution of as H(μ) which represents the probability that the value of is
equal or lower than a certain value μ i.e.
{ } (3)
The probability distribution H(μ) depends on plant characteristics and the functional relationship is different for
wind and solar.
A. Wind Variability Distribution
The functional form of H(μ) defined in (3) for wind power generation approximately follows the empirical relation
[19],
(4)
Here K(Δt) is a model parameter which depends on the plant characteristics and can have seasonal variation. The
wind power actual generation and the distribution are shown in fig. 1.

(a) (b)
(c)
Fig 1. (a) shows the wind power data W(t) for the state of Karnataka, India from July, 2010 to September, 2011 and (b) shows
the of W(t) for Δt = 5. (c) shows the wind power distribution H(μ). The horizontal axis represents the values of μ and the
vertical axis shows the probability values of H(μ).
B. Solar Variability Distribution
For solar power generation, the functional form of H(μ) defined in (3) approximately follows the empirical relation
as,
{ } if

{ } if (5)
The solar power actual generation and the distribution are shown in fig. 2.
(a) (b)
Fig 2: (a) Solar power output of two different days and (b) of respective days and (c) shows H(μ) of two different
days. The dotted marker shows the actual data and the lines show the distribution using (4)
III. ACCURACY OF FORECASTING & SCHEDULING AND PENALTY DUE TO DEVIATION
Due to variability and uncertainty, forecasting is an important aid to effective and efficient planning; The
forecasting and scheduling strategy in variable renewable energy is an important factor in grid stability for high
penetration of wind and solar energy. Though the scheduling is mandatory with effect from January 1, 2012 in
India, earlier this year, the Central Electricity Regulatory Commission (CERC) introduced a robust framework to

strengthen forecasting in the renewable sector in India. The CERC also issued the Indian Electricity Grid Code
(Third Amendment) Regulations, 2015 and Deviation Settlement Mechanism and related matters (Second
Amendment), Regulations 2015. Some useful features of the mechanism are :
 The mechanism shall be applicable to wind and solar generators.
 The maximum number of revisions has been increased from 8 to 16.
 The penalties for deviation have been computed as per Power Purchase Agreements and shall be levied
for any deviation beyond +/-15%
According to the CERC Interstate mechanism notified vide notification no. 1/14/2015-Reg.Aff.(FSDS)(ii)/CERC
Dated: 07.08.2015 and F.O.R Model regulation dated 05.11.2015 by forum of regulators, the forecasting error
can be defined as
(6)
According to regulation the generalised structure penalty due to deviation can be represented as follows:
Table I: Generalized Structure of Deviation charge
Error Band Deviation Charge
per kw-Hr (Rs)
PPA Based Fixed
| | No penalty 0
| | 10% of PPA INR 0.50
| | 20% of PPA INR 1.00
| | 30% of PPA INR 1.50
Here the value of m, m1 and m2 takes different value for different regulations and can differ for wind and solar or
can differ for existing and new commissioned projects, but interestingly the following equation holds for all existing
regulations,
(7)
For example for CERC regulation, m = 15%, m1 = 25% and m2 = 25% for wind and solar.
A. Accuracy of Forecasting & Scheduling
Considering the error defined in (6), the temporal accuracy of forecasting for each day can be calculated as

(8)
where m takes value 15% for CERC and can differ for other regulations.
Unlike temporal behaviour, the energy based performance criteria can be defined as:
Net Accuracy in Energy = | | (9.A)
Deviated Energy in % = ∑ (9.B)
Where,
R1 = Range1: Deviation in between m % to m1 % of AvC.
R2 = Range2: Deviation in between m1 % to m2 % of AvC.
R3 = Range3: Deviation greater than m2 % of AvC.
B. Penalty due to Deviation
It is required to calculate the approximate average cost of penalties for deviation of wind power forecasting.
Without showing the detailed statistical analysis, this article shows some important statistical relations and
approximations to measure the cost of penalty. For simplification let consider,
C = average cost per available capacity
c(e) = cost of penalty due to error e where error in new regulation is defined as , here AvC =
available capacity, xa = actual power and xf = forecast power
If h(e) represents the probability distribution of error e, it is easy to show that the cost per available capacity can
be represented as,
∫ (10)
For a good forecasting with maximum 16 revision, we can consider that the mean of distribution h(e) is
approximately 0, variance is and h(- e) = h(e). Considering the no penalty band as [-m,+m], we can consider
the deviation charge follows an linear relation as
| | if | | (11)
= 0 , otherwise
Using some algebraic manipulation we can show that
[ ∫ ] (12)
The integral part of the right hand side of the previous equation can be approximated and with some basic
approximation we can state that

[ ] (13)
Where
∫ (14)
Here represents the probability that the absolute error |e| lies in [0, m]. Considering the Table II, we can
state that
(15)
Hence, the average cost per available capacity can be represented as,
[ ] (16)
Table II: Error band and slab wise Penalty for given Probability Values
Deviation Charge
per kw-Hr (Rs)
Error Band Mean Value Probability Slab wise Penalty
PPA Based Fixed
No penalty 0 | | NA
10% of PPA INR 0.50 | |
20% of PPA INR 1.00 | |
30% of PPA INR 1.50 | |
IV. FORECASTING & SCHEDULING SOLUTION OF WIND AND SOLAR POWER BY
Due to stochastic behavior of wind and solar power generation and the variability distribution of power, the
prediction of ramping patterns creates uncertainties in forecasting. Though numerous models are available using
different methodology, a good forecast captures the genuine patterns which exist in the historical data, but do not
replicate past events that will not occur again.
A. Forecasting & Scheduling Solution
The del2infinity Forecasting System is developed considering India-specific situations where access to possible

and necessary wind/solar power prediction data is a major issue. The system has been created considering the
non-availability of some parameter data in many cases. However, the system is capable of accommodating and
optimizing on additional data when available. We believe in using the available data intelligently instead of asking
for more data that may not be available.
We uses Power to Power (P2P) algorithm as the basic functional module of the analytics engine. In this module
the input is only the wind/solar-power time series data (aggregated or plant level). The analytics engine uses
different statistical pattern matching and ANN algorithms to fetch different statistical information related to the
wind/solar power data. The P2P algorithm generates different patterns in wind/solar power data as well as
information on wind/solar power ramping events. The system forecasts the next day‟s wind/solar power output in
15 minute time blocks. The P2P algorithm does not use wind speed/radiation (like DNI) or weather related data to
ensure system performance when there is lack of such data.
The power to power („P2P‟) forecast methodology developed by us is used to forecast the solar and wind power
data points. The P2P forecast methodology works better if there is no curtailment issue. But the curtailment issue
of one day creates erroneous forecast in next day and affects next few days if only P2P Artificial Intelligence („AI‟)
system is used. Hence though the proposed methodology uses P2P AI system, two parallel weather feedback
loops are introduced to adapt the system. One feedback loop is used for pattern matching using support vector
machine („SVM‟) and the other loop is used to measure the stochastic variation of power input. The analytics
engine also has a number of feedback loops that use wind/solar and weather data. Data used in these feedback
loops can include numerical weather prediction variables (e.g. forecasted wind speed, direction, and pressure /
DNI, ambient temperature etc) and weather observations. These parallel feedback loops are activated only when
these data are available.
B. Results
The Forecast and scheduling system is used for generation power prediction of different solar and wind power
projects and shows encouraging results minimizing the penalty due to deviation mechanism. As shown in figure
3(a) the forecast of wind power without any revision shows satisfactory results for 103 MW plant in India. The
accuracy can be further increased with multiple revisions. Figure 3(b) shows an outcome of the forecasting
system of a solar plant of capacity 40 MW.
(a)

(b)
Fig 3: Actual and schedule generation of (a) wind power plant and (b) solar power plant in India
C. Penalty Analysis
The monthly penalty due to deviation is analyzed for different wind and solar plants. A sample analysis for wind
and solar plants are shown here in Table III and IV for an easy understanding.

Table III: Monthly Penalty of 27 MW Wind Plant in India
Table IV: Monthly Penalty of 10 MW Solar Plant in India
V. CONCLUSION
Forecasting and scheduling is an essential requirement for grid stability and a sustainable energy future of India.
This paper describes the result oriented forecasting and scheduling solution which optimizes the penalty due to
deviation considering the existing regulations. The variability of wind and solar energy is briefly described to
understand the necessity of forecasting; and a computational framework of penalty due to deviation is analyzed in
this paper. Some examples and analysis regarding the actual and scheduled generation of power for wind and
solar plants in India are presented showing the feasibility of forecasting and scheduling regulation of India with
satisfactory level of accuracy. Though there is a paucity of historical data and the data related to weather
parameters of different solar and wind plants in India, but our system shows encouraging results and will be
evolved in future with better accuracy.

REFERENCES
[1] http://timesofindia.indiatimes.com/india/Electricity-for-18000-villages-in-1000-days-PM-Modi/articleshow/49774732.cms last visited
on 30 Sept. 2016
[2] National Action Plan on Climate Change, GOI (http://www.moef.nic.in/downloads/home/Pg01-52.pdf last visited on 30 Sept. 2016)
[3] INDIA‟S INTENDED NATIONALLY DETERMINED CONTRIBUTION: WORKING TOWARDS CLIMATE JUSTICE
(http://www4.unfccc.int/submissions/INDC/Published%20Documents/India/1/INDIA%20INDC%20TO%20UNFCCC.pdf last visited
on 30 Sept. 2016)
[4] Report of the Expert Group on 175 GW RE by 2022, NITI Aayog, GOI
(http://niti.gov.in/writereaddata/files/writereaddata/files/document_publication/report-175-GW-RE.pdf last visited on 30 Sept. 2016)
[5] Strategic Plan for new and renewable Energy Sector for the period 2011-17, Ministry of New and Renewable Energy, Government
of India, 2011
(http://mnre.gov.in/file-manager/UserFiles/strategic_plan_mnre_2011_17.pdf last visited on 30 Sept. 2016)
[6] Das Abhik Kumar, “An Empirical Model of Power Curve of a Wind Turbine”, Energy Systems vol. 5(3), pp. 507-518, March 2014
[7] Tai-Her, Y., Li, W.: A study on generator capacity for wind turbines under various tower heights and rated wind speeds using
Weibull distribution. IEEE Trans. Energy Convers. 23, 592–602 (2008)
[8] Mathew, S., Philip, G.S.: Advances in Wind Energy Conversion Technology. Springer, New York (2011)
[9] Das Abhik Kumar , "Analytical Derivation of Explicit J–V Model of a Solar Cell from Physics based Implicit Model", Solar Energy,
Elsevier, vol. 86, issue 1, pp 26-30, January 2012
[10] Das Abhik Kumar , "An Explicit J–V Model of a Solar Cell for Simple Fill Factor Calculation", Solar Energy, Elsevier, vol. 85, issue 9,
pp 1906-1909, September 2011
[11] Das Abhik Kumar , Karmalkar Shreepad, "Analytical Derivation of the Closed-form Power Law J-V Model of an lluminated Solar Cell
from the Physics Based Implicit Model", IEEE Transactions on Electron Devices, vol. 58, No 4, pp 1176-1181, April 2011
[12] Das Abhik Kumar, “An analytical model for ratio based analysis of wind power ramp events”, Sustainable Energy Technology and
Assessments, Elsevier vol. 9, pp.49-54, March 2015
[13] Mazumdar Bishal Madhab, Md. Saquib, Das Abhik Kumar “An Empirical Model for Ramp Analysis of Utility-Scale Solar PV Power”, ,
Solar Energy, Elsevier, vol. 107, pp. 44-49, September 2014
[14] Indian Electricity Grid Code, Central Electricity Regulatory Commission, 2010
(http://cercind.gov.in/2010/ORDER/February2010/IEGC_Review_Proposal.pdf last visited on 30 Sept. 2016)
[15] Procedure for implementation of the mechanism of Renewable Regulatory Fund , Central Electricity Regulatory Commission, 2011
( http://www.cercind.gov.in/Regulations/Detailed_Procedure_IEGC.pdf last visited on 30 Sept. 2016)
[16] Framework on Forecasting, Scheduling and Imbalance Handling for Variable Renewable Energy Sources (Wind and Solar), Central
Electricity Regulatory Commission,
(http://www.cercind.gov.in/2015/regulation/SOR7.pdf last visited on 30 Sept. 2016)
[17] Model Regulations on Forecasting,Scheduling and Deviation Settlement of Wind and Solar Generating Stations at the State level
(http://www.forumofregulators.gov.in/Data/study/MR.pdf last visited on 30 Sept. 2016 )
[18] Das Abhik Kumar, “Quantifying photovoltaic power variability using Lorenz curve”, Journal of Renewable and Sustainable Energy,
AIP, vol.6 (3), June 2014
[19] Kamath, C. 2010. “Understanding Wind Ramp Events through Analysis of Historical Data.” Transmission and Distribution
Conference and Exposition, 2010 IEEE PES in New Orleans, LA, United States., April 2010
[20] Das Abhik Kumar & Majumder Bishal Madhab, “Statistical Model for Wind Power based on Ramp Analysis”, International Journal
of Green Energy, 2013
[21] Gallego C., Costa A., Cuerva A., Landberg L., Greaves B., Collins J., “A wavelet-based approach for large wind power ramp
characterisation”, Wind Energy, vol. 16(2), pp. 257-278, Mar. 2013
[22] Bosavy A., Girad R., Kariniotakis G., “Forecasting ramps of wind power production with numerical weather prediction ensembles”,
Wind Energy, vol. 16(1), pp. 51-63, Jan. 2013
[23] Kirby B., Milligan M., “An exemption of capacity and ramping impacts of wind energy on power systems”, The Electricity Journal,
vol.2(7), Sept. 2008, pp.30-42

Forecasting and scheduling of wind and solar power generation in india

More Related Content

What's hot (17)

Similar to Forecasting and scheduling of wind and solar power generation in india (20)

More from Das A. K. (20)

Recently uploaded (20)

Forecasting and scheduling of wind and solar power generation in india