180625 Antares General Reference Guide

Antares-general-reference-guide

User Manual:

Open the PDF directly: View PDF .
Page Count: 57

Download
Open PDF In Browser	View PDF

Antares_Simulator 6.0.0
GENERAL REFERENCE GUIDE

Simulation package
Script Editor package
Graph Editor package
Data Organizer package

X

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

1/57

ANTARES QUICK REFERENCE GUIDE
Table of contents
1 Introduction............................................................................................................................................................................. 3
2 General content of Antares sessions ....................................................................................................................................... 3
2 Data organization .................................................................................................................................................................... 5
3 Commands............................................................................................................................................................................... 7
File ......................................................................................................................................................................................... 7
Edit ........................................................................................................................................................................................ 8
Input ...................................................................................................................................................................................... 8
Output ................................................................................................................................................................................... 9
Run ...................................................................................................................................................................................... 10
Configure............................................................................................................................................................................. 10
Scripts.................................................................................................................................................................................. 12
Tools.................................................................................................................................................................................... 12
Window ............................................................................................................................................................................... 13
“?” Menu ............................................................................................................................................................................. 13
4 Active windows ..................................................................................................................................................................... 14
System Maps ....................................................................................................................................................................... 14
Simulation ........................................................................................................................................................................... 14
User’s Notes ........................................................................................................................................................................ 18
Load..................................................................................................................................................................................... 18
Thermal ............................................................................................................................................................................... 19
Hydro .................................................................................................................................................................................. 21
Wind .................................................................................................................................................................................... 25
Solar .................................................................................................................................................................................... 26
Misc. Gen. ........................................................................................................................................................................... 27
Reserves / DSM ................................................................................................................................................................... 28
Links .................................................................................................................................................................................... 28
Binding constraints ............................................................................................................................................................. 29
Economic Opt. ..................................................................................................................................................................... 30
Miscellaneous ..................................................................................................................................................................... 30
5 Output files ............................................................................................................................................................................ 31
Economy and Adequacy, area results ................................................................................................................................. 32
Economy and Adequacy, interconnection results ............................................................................................................... 33
Economy and Adequacy, other results ............................................................................................................................... 34
Draft, area results ............................................................................................................................................................... 34
Miscellaneous ..................................................................................................................................................................... 35
6 Time-series analysis and generation ..................................................................................................................................... 36
General................................................................................................................................................................................ 36
Time-series generation (load, wind, solar) : principles ...................................................................................................... 37
Time-series generation (load, wind, solar) : GUI ................................................................................................................ 38
Time-series analysis (load, wind, solar) ............................................................................................................................... 40
Time-series generation (thermal) ....................................................................................................................................... 42
Time-series analysis (thermal) ............................................................................................................................................ 46
Time-series generation and analysis (hydro) ...................................................................................................................... 47
7 Miscellaneous ........................................................................................................................................................................ 48
Antares at one glance ......................................................................................................................................................... 48
Operating reserves modeling .............................................................................................................................................. 49
Conventions regarding colors and names ........................................................................................................................... 51
Definition of geographic districts ........................................................................................................................................ 52
The “export mps” optimization preference ........................................................................................................................ 54
8 System requirements ............................................................................................................................................................ 55
Operating system ................................................................................................................................................................ 55
Hard drive disk .................................................................................................................................................................... 55
Memory .............................................................................................................................................................................. 55
Multi-threading ................................................................................................................................................................... 55

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

2/57

1 Introduction
This document describes all of the main features of the Antares_Simulator package, version 6.0.0.
It gives useful general information regarding the way data are handled and processed, as well as how
the Graphic User Interface (GUI) works. So as to keep this documentation as compact as possible,
many redundant details (how to mouse-select, etc.) are omitted.
Some features described in this guide are not fully operational in 6.0.0 version. Features not yet
available appear in grey in the GUI.

Real-life use of the software involves a learning curve process that cannot be supported by a simple
reference guide. So as to be able to address this basic issue, two kinds of resources may be used:

The examples library, which is meant as a self-teaching way to learn how to use the software.
It is enhanced in parallel to the development of new features. The content of this library may
depend on the type of installation package it comes from (general public or members of the
users’ club).

The https://antares.rte-france.com website

Please report misprints or other errors to:
Rte-antares@rte-france.com

2 General content of Antares sessions
A typical Antares session involves different steps that are usually run in sequence, either automatically
or with some degree of man-in-the-loop control, depending on the kind of study to perform.
These steps most often involve:
a) GUI session dedicated to the initialization or to the updating of various input data sections
(load time-series, grid topology, wind speed probability distribution, etc.)
b) GUI session dedicated to the definition of simulation contexts (definition of the number and
consistency of the “Monte-Carlo years” to simulate)
c) Simulation session producing actual numeric scenarios following the directives defined in (b)
d) Optimization session aiming at solving all of the optimization problems associated with each of
the scenarios produced in (c).
e) GUI session dedicated to the exploitation of the detailed results yielded by (d)
The scope of this document is to describe the features of the software involved in step (a) to (e), from
the user’s standpoint1.
The following picture gives a functional view of all that is involved in steps (a) to (e).

1
This guide does not provide a detailed expression of the mathematical problems solved. Note, however, that a comprehensive
formulation of all optimization problems may be printed along with the simulation results if the user so requires.

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

3/57

The number and the size of the individual problems to solve (typically, a least-cost hydro-thermal
power schedule and unit commitment, with an hourly resolution and throughout a week, over a large
interconnected system) make optimization sessions often computer-intensive.
Depending on user-defined results accuracy requirements, various practical options allow to simplify
either the formulation of the problems or their resolution.
In terms of power studies, the different fields of application Antares has been designed for are the
following:
Generation adequacy problems : assessment of the need for new generating plants so as to
keep the security of supply above a given critical threshold
What is most important in these studies is to survey a great number of scenarios that
represent well enough the random factors that may affect the balance between load and
generation. Economic parameters do not play as much a critical role as they do in the
other kinds of studies since the stakes are mainly to know if and when supply security is
likely to be jeopardized (detailed costs incurred in more ordinary conditions are of
comparatively lower importance). In these studies, the default Antares option to use is the
“Adequacy” simulation mode, or the “Draft” simulation mode (which is extremely fast but
which produces crude results).
Transmission project profitability : assessment of the savings brought by a specific
reinforcement of the grid, in terms of decrease of the overall system generation cost (using an
assumption of fair and perfect market) and/or improvement of the security of supply (reduction
of the loss-of-load expectation).
In these studies, economic parameters and the physical modeling of the dynamic
constraints bearing on the generating units are of paramount importance. Though a
thorough survey of many “Monte-Carlo years” is still required, the number of scenarios to
simulate is not as large as in generation adequacy studies. In these studies, the default
Antares option to use is the “Economy” simulation mode.
The common rationale of the modeling used in all of these studies is, whenever it is possible, to
decompose the general issue (representation of the system behavior throughout many years, with a
time step of one hour) into a series of standardized smaller problems.
In Antares, the “elementary“ optimization problem resulting from this approach is that of the
minimization of the overall system operation cost over a week, taking into account all proportional and
non-proportional generation costs, as well as transmission charges and “external” costs such as that
of the unsupplied energy (generation shortage) or, conversely, that of the spilled energy (generation
excess). In this light, carrying out generation adequacy studies or transmission projects studies means
formulating and solving a series of a great many week-long operation problems (one for each week of
each Monte-Carlo year ), assumed to be independent (note that, however, issues such as the
management of hydro resources may bring some degree of coupling between the successive
problems).
Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

4/57

2 Data organization
In Antares, all input and output data regarding a given study are located within a folder named after
the study and which should preferably be stored within a dedicated library of studies (for instance:
C/.../A_name_for_an_Antares_lib/Study-number-one).
The software has been designed so that all input data may be handled (initialized, updated, deleted)
through the simulator’s GUI. Likewise, all results in the output data can be displayed and analyzed
within the simulator: its standard GUI is actually meant to be able to provide, on a stand-alone basis,
all the features required to access input and output in a user-friendly way.
In addition to that, the Antares 6.x simulator may come2 with or without functional extensions that
provide additional ways to handle input and output data.
These extensions take the form of companion applications whose documentation is independent from
that of the main simulator. For information regarding these tools (Graph Editor, Script Editor, Study
Manager) please refer to the specific relevant documents.
Besides, a point of notice is that most of Antares files belong to either “.txt” or “.csv” type: as an
alternative to the standard GUI, they can therefore be viewed and updated by many applications
(Windows Notepad, Excel,….). However, this is not recommended since handling data this way may
result in fatal data corruption (e.g. as a consequence of accidental insertion of special characters).
Direct access to input or output data files should therefore be reserved to very experienced users.
The input data contained in the study folder describe the whole state of development of the
interconnected power system (namely: grid, load and generating plants of every kind) for a given
future year.
As already stated, all of these data may be reviewed, updated, deleted through the GUI, whose
commands and windows are described in Sections 3 and 4.
Once the input data is ready for calculation purposes, an Antares session may start and involve any or
all of the following steps: historical time-series analysis, stochastic times-series generation, (draft)
adequacy simulation, (full) adequacy simulation and economic simulation.
The results of the session are stored within the output section of the study folder. The results obtained
in the different sessions are stored side by side and tagged. The identification tag has two
components: a user-defined session name and the time at which the session was launched.

2

Depending on the installation package

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

5/57

Particular cases are:
a) The outputs of the Antares time-series analyzer are not printed in the general output files but
kept within the input files structure (the reason being that they are input data for the Antares
time - series generators). The associated data may nonetheless be accessed to be reviewed,
updated and deleted at any time through the GUI.
b) Some specific input data may be located outside the study folder: this is the case of the
historical times-series to be processed by the time-series analyzer, which may be stored either
within the “user” subfolder of the study or anywhere else (for instance, on a remote “historical
data” or “Meteorological data” server).
c) The study folder contains a specific subfolder named “user”, whose status is particular:
Antares is not allowed to delete any files within it (yet files may be updated on the user’s
requirement). As a consequence, the “user” subfolder is unaffected by the “clean study”
command (see Section 3). This subfolder is therefore a “private” user space, where all kinds of
information can be stored without any kind of interference with Antares. Note that on using the
“save as” command (described further below), the choice is given to make or not a copy of this
subfolder.

d) The times-series analyzer requires the use of a temporary directory in which intermediate files
are created in the course of the analysis and deleted in the end, unless the user wishes to
keep them for further examination. Its location is user-defined and should usually be the “user”
subfolder if files are to be kept, otherwise any proper temporary space such as “C..../Temp”.

e) If the interconnected system to study is large and/or if the computer is low on RAM, it is
possible to run the Monte-Carlo adequacy simulator as well as the Monte-Carlo economic
simulator in “Swap” mode. Swap is not handled by the computer’s OS but by an Antares
specific swap manager, whose operation requires the definition of a space where the software
can store temporary files. This location is user-defined but should never be chosen within the
study folder. C/.../Temp may typically be used but an external drive may be preferred if the
computer is low on HDD.

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

6/57

3 Commands
The Antares GUI gives access to a general menu of commands whose name and meanings are
described hereafter.

File
New
Creates a new empty study to be defined entirely from scratch (network
topology, interconnections ratings, thermal power plants list, fuel costs, hydro
inflows stats, wind speed stats, load profiles ,etc.)
Open
Loads in memory data located in a specified Antares study folder. Once
loaded, these data may be reviewed, updated, deleted, and simulations may
be performed. If “open” is performed while a study was already opened, the
former study will be automatically closed.
Quick Open
Same action as open, with a direct access to the recently opened studies
Save
Saves the current state of the study, if necessary by replacing original files by
updated ones. After using this command the original study is no longer
available, though some original files may be kept until the ”clean” command is
used (see “clean” command )
Save as
Saves the current state of the study under a different name and / or location.
Using this command does not affect the original study. When “saving as”, the
user may choose whether he prefers to save input and output data or only
input data. Note that Antares does not perform “autosave”: Therefore, the
actions performed on the input data during an Antares session (adding an
interconnection, removing a plant,...) will have no effect until either “save” or
“save as” have been used
Export Map
Saves a picture of the current map as a PNG, JPEG or SVG file
Default background colour and storage location can be changed

Open in Windows Explorer
Opens the folder containing the study in a standard Windows Explorer window
Clean
Removes all junk files that may remain in the study folder if the Antares
session has involved lots of sequences such as “create area – add plant –
save –rename area – save - rename plant ...” (Antares performs only low level
auto-clean for the sake of GUI’s efficiency)
Close
Closes the study folder. If no “save” or “save as” commands have been
performed, all the modifications made on the input data during the Antares
session will be ignored
Quit
Exits from Antares

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

7/57

Edit
Copy
Prepare a copy of elements selected on the current system map. The
command is available only if the current active tab (whose name appears at
the top line of the subcommand menu) is actually that of the System maps.
Paste
Paste elements previously prepared for copy. The command is available only
if the current active tab (whose name appears at the top line of the
subcommand menu) is actually that of the System maps. Note that
copy/paste may be performed either within the same map or between two
different maps, attached to the same study or to different studies. To achieve
that, launch one instance of Antares to open the “origin” study, select
elements on the map and perform copy, launch another instance of Antares to
open the destination study, perform paste. Copied elements are stored in an
Antares clipboard that remains available for subsequent (multiple) paste as
long as the system map is used as active window.
Paste Special
Same as Paste, with a comprehensive set of parameterized actions (skip,
merge, update, import) that can be defined for each data cluster copied in the
clipboard. This gives a high level of flexibility for carrying out complex
copy/paste actions.
Reverse
The elements currently selected on the system map are no longer selected
and are replaced by those not selected beforehand.
Unselect All
Unselect all elements currently selected on the system map.
Select All
Select all elements on the system map.

Input
Name of the study
Gives a reference name to the study. The default name is identical to that of
the study’s folder but the user may modify it. The default name of a new study
is “no title”
Author(s)
Sets the study’s author(s) name. Default value is “memory”
The other “input” subcommands here below are used to move from one active window to another

System Maps
Simulation
User’s Notes
Load
Solar
Wind
Hydro
Thermal
Misc. Gen.
Reserves/DSM
Links
Binding constraints
Economic opt.

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

8/57

Output
 < simulation tag>
For each simulation run for which results have been generated, opens a GUI
for displaying results. Results may be viewed by multiple selections made on a
number of parameters. Note that, since all simulations do not include all kinds
of results (depending on user’s choices), some parameters are not always
visible. Parameters stand as follows:

•

Antares area (node)

•

Antares interconnection (link)

•

Class of Monte-Carlo results :
o Monte-Carlo synthesis (over all years simulated)
o Year-by-Year (detailed results for one specific year)

•

Category of Monte-Carlo results :
o General values (operating cost, generation breakdown, ...)
o Thermal plants (detailed thermal generation breakdown)
o Record years (for each Antares variable, identification of
the Monte-Carlo year for which lowest and highest values
were encountered)

•

Span of Monte-Carlo results :
o
o
o
o
o

Hourly
Daily
Weekly
Monthly
Annual

The interface provides a user-friendly way for the comparison of results between multiple
simulations (e.g. “before” and “after” commissioning of a new plant or interconnection) :

•
•

Use “new tab” button and choose a first set of simulation results
Use again “new tab” and choose a second set of simulation results

The results window will be automatically split so as to show the two series of results in parallel.
To the right of the “new tab” button, a symbolic (icon) button gives further means to compare
results on a split window (average, differences, minimum, maximum, sum)
Besides, when the simulation results contain the “year-by-year” class, it is possible to carry out
an extraction query on any given specific variable (e.g. “monthly amounts of CO2 tons
emitted”) throughout all available years of simulation.
The results of such queries are automatically stored within the output file structures, so as to
be available at very short notice if they have to be examined later in another session
(extractions may require a significant computer time when there are many Monte-Carlo years
to process).
Open in Windows Explorer
Displays the list of available simulation results and allows browsing through
the output files structure. The content of these files may be reviewed by tools
such as Xcel. File structures are detailed in Section 5.

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

9/57

Run
Monte Carlo Simulation
Runs either an economy simulation, an adequacy simulation, or a “draft”
simulation, depending on the values of the parameters set in the “simulation”
active window (see Section 4). If hardware resources and simulation settings
allow it, simulations benefit from full multi-threading (see Section 8)
Time-series generators
Runs any or all of the Antares stochastic time-series generators, depending
on the values of the parameters set in the “simulation” active window (see
Section 6)
Time-series analyzer
Runs the Antares historical time-series analyzer. The parameters of this
module are defined by a specific active window, available only on launching
the analyzer (see Section 6)

Configure
Filters on simulation results
Opens a few auxiliary windows that allow multiple selection on the results to
store at the end of a simulation: Choice of areas or geographic districts (see
below), choice of interconnections, choice of results spans (hourly, daily, etc.).
Note that in versions where the feature is not operational (grey display), an
alternative way of filtering results is available (see Section 4, “output profile”).
Geographic Districts
Allows selecting a set of areas so as to bundle them together in a “district”.
These are used in the course of simulations to aggregate results over several
areas. They can be given almost any name (a “@” prefix is automatically
added by Antares). Bypassing the GUI is possible (see Section 7).
MC Scenario builder
For each Monte-Carlo year of the simulation defined in the “Simulation”
window, this command allows to state, for each kind of time-series, whether it
should be randomly drawn from the available set (be it ready-made or
Antares-generated) OR should take a user-defined value (in the former case,
the default “rand” value should be kept; in the latter, the value should be the
reference number of the time-series to use). Multiple simulation profiles can
be defined and archived. The default active profile gives the “rand” status for
all time-series in all areas (full probabilistic simulation).
MC Scenario playlist
For each Monte-Carlo year of the simulation defined in the “Simulation” active
window, this command allows to state whether a MC year prepared for the
simulation should be actually simulated or not. This feature allows, for
instance, to refine a previous simulation by excluding a small number of “raw”
MC years whose detailed analysis may have shown that they were not
physically realistic. A different typical use consists in replaying only a small
number of years of specific interest (for instance , years in the course of which
Min or Max values of a given variable were encountered in a previous
simulation).
Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

10/57

Object custom properties
Opens an interface that allows multiple selection of Antares objects (nodes,
links, thermal clusters,etc.) and to bind specific data to them (e.g. minimum
voltage level, maximum voltage level, etc ...).
These data are not used by Antares itself but may prove useful for external
applications

Optimization preferences
Defines a set of options related to the optimization core used in the
simulations. The set of preferences is study-specific; it can be changed at any
time and saved along with study data. Options refer to objects (binding
constraints,tec.) that are presented in subsequent sections of this document.
The values set in this menu overlay the local parameters but do not change
their value: for instance, if the LOCAL parameter “set to infinite” is activated for
some interconnections, and if the GLOBAL preference regarding transmission
capacities is “set to null”, the simulation will be carried out as if there were no
longer any grid BUT the local values will remain untouched. If the preference
is afterwards set to “local values”, the interconnections will be given back their
regular capacities (infinite for those being set on “set to infinite”).

•

Binding constraints (include / ignore)

•

Hurdle costs (include / ignore)

•

Transmission capacities (local values / set to null / set to infinite)

•

Min Up/down time of thermal plants (include / ignore)

•

Day-ahead reserve (include / ignore)

•

Primary reserve (include / ignore)

•

Strategic reserve (include / ignore)

•

Spinning reserve (include / ignore)

•

Export mps (false/true) (see Sections 7 and 8)

•

Simplex optimization range3 (day / week)

3

Weekly optimization performs a more refined unit commitment, especially when the level selected in the “advanced
parameters” menu is “accurate”.

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

11/57

Advanced parameters
These parameters seldom need to be changed. The set of parameters is
study-specific; it can be updated at any time.

•

Seeds for random number generation
o Time-series draws (MC scenario builder)
o Wind time-series generation
o Solar time-series generation
o Hydro time - series generation
o Load time - series generation
o Thermal time-series generation
o Noise on thermal plants costs
o Noise on unsupplied energy costs
o Noise on spilled energy costs
o Noise on virtual hydro cost
o Initial hydro reservoir levels
• Spatial time-series correlation
o Numeric Quality : load [standard | high]
o Numeric Quality : wind[standard | high]
o Numeric Quality : solar[standard | high]
• Other preferences
o Power fluctuations [free modulations | minimize
excursions | minimize ramping]
o Shedding policy [shave peaks | minimize duration]
o District marginal prices : [average | weighed]
o Day-ahead reserve management [global|local]
o Unit commitment mode [fast |accurate]
o Simulation cores [minimum|low|medium|high|maximum]

Scripts

Allows execution of R- scripts edited through the “script editor” package (see
dedicated documentation for this package)

Tools
Study manager
Launches the “study manager” external package
(Please refer to dedicated documentation for this package)
Grapher
Launches the “graph editor” external package
(Please refer to dedicated documentation for this package)
CSV viewer
Opens txt or csv files, with a set of minimal data handling functions
(copy/paste, find min, max, compute average, standard deviation, etc.)
Resources monitor
Indicates the amounts of RAM and disk space currently used and those
required for a simulation in the available modes (see Section 8). Note that the
“disk requirement” amount does not include the footprint of the specific “mps”
files that may have to be written aside from the regular output (see previous §
“optimisation preferences”). Besides, the resources monitor shows the number
of CPU cores available on the machine Antares is running on.
Configure the swap folder
Defines the location that will be used by Antares to store the temporary files of
the MC simulators when the swap mode is activated (this location may also be
used by Antares GUI when handling large studies). The default setting is the
system temporary folder
Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

12/57

Window
Toggle full window
Uses the whole window for display
Inspector
Opens a window that gives general information on the study and allows quick
browsing through various area- or interconnection-related parameters
Log viewer
Displays the log files regarding every Antares session performed on the study

“?” Menu
Reference Guide
Short-cut to this document (pdf reader software required)
System Map Editor Reference Guide
Short-cut to a guide to copy/paste features (pdf reader software required)
Continue on-line
Connects to the internet (required to participate to anonymous usage metrics
and, more generally, to get access to other services - future versions).
Privacy Policy – GDPR compliance:
When an Antares_Simulator GUI connects to the internet, it sends a signal to
a server dedicated to the gathering of anonymous usage metrics, over a
rolling period of one year.
The communication
does not convey any personal data: transmitted
information is limited to three items:
•
•
•

Antares_Simulator version number
Computing power range of the machine running Antares_Simulator
(number of CPU cores)
Signature of the running instance of Antares_Simulator

Continue off-line
Disconnects from the internet. All anonymous usage metrics that may have
been gathered so far for this instance of Antares_Simulator will be discarded.
Show signature
Displays the anonymous signature under which this instance of
Antares_Simulator will be referred to in web-based services, if it goes on-line

Check for updates
Tells if a more recent version of Antares_Simulator is available

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

13/57

4 Active windows
Data can be reviewed, updated, deleted by selecting different possible active windows whose list and
content are described hereafter. On launching Antares, the default active window is “System Maps”.

System Maps
This window is used to define the general structure of the system, i.e. the list of areas and that
of the interconnections. Only the area’s names, location and the topology of the grid are
defined at this stage. Different colours may be assigned to different areas. These colours may
later be used as sorting options in most windows. They are useful to edit data in a fashion that
has a geographic meaning (which the lexicographic order may not have).
This window displays copy/paste/select_all icons equivalent to the relevant EDIT menu
commands.
The top left side of the window shows a “mouse status” field with three icons. These icons
(one for nodes, one for links and one for binding constraints) indicate whether a selection
made on the map with the mouse will involve or not the related elements.
When a copy/paste action is considered, this allows for instance to copy any combination of
nodes, links and binding constraints. Status can be changed by toggling the icons. Default is
“on” for the three icons.
Two other purely graphic icons/buttons (no action on data) allow respectively to centre the
map on a given set of (x,y) coordinates, and to prune the “empty” space around the current
map.
Multiple additional maps may be defined by using the cross-shaped button located top right. A
detailed presentation of all system map editor features can be found in the document “System
Map Editor Reference Guide”.

Simulation
The main simulation window is divided up in two parts. On the left side are the general
parameters while the right side is devoted to the time-series management. These two parts
are detailed hereafter

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

14/57

LEFT PART : General parameters
Simulation
Mode:

Economy , Adequacy, Draft4

First day:

First day of the simulation (e.g. 8 for a simulation beginning on
the second week of the first month of the year)

Last day:

Last day of the simulation (e.g. 28 for a simulation ending on
the fourth week of the first month of the year)5

Calendar
Horizon:

Reference year (static tag, not used in the calculations)

Year:

Actual month by which the Time-series begin (Jan to Dec, Oct
to Sep, etc.)

Leap Year:

(Yes/No) indicates whether February has 28 or 29 days

Week:

In economy or adequacy simulations, indicates the frame
(Mon- Sun, Sat-Fri, etc.) to use for the edition of weekly
results

1st January:

First day of the year (Mon, Tue, etc.)

Monte-Carlo scenarios
Number:

Number of MC years that should be prepared for the
simulation (not always the same as the Number of MC years
actually simulated, see “selection mode” below)

Building mode:
(Automatic)
For all years to simulate, all time-series will be drawn at
random
(Custom)
The simulation will be carried out on a mix of deterministic and
probabilistic conditions, with some time-series randomly drawn
and others set to user-defined values. This option allows
setting up detailed “what if” simulations that may help to
understand the phenomena at work and quantify various kinds
of risk indicators. To set up the simulation profile, choose in
the main menu: Configure/ MC scenario builder

4
”Economy” simulations make a full use of Antares optimization capabilities. They require economic as well as technical input
data and may demand a lot of computer resources. “Adequacy” simulations are faster and require only technical input data.
Their results are limited to adequacy indicators. “Draft” simulations are highly simplified adequacy simulations, in which binding
constraints (e.g. DC flow rules) are ignored, while hydro storage is assumed to be able to provide its nominal maximum power
whenever needed. As a consequence, draft simulations are biased towards optimism. They are, however, much faster than
adequacy and economic simulations.
5
In Economy an Adequacy simulations, these should be chosen so as to make the simulation span a round number of weeks.
If not, the simulation span will be truncated: for instance, (1, 365) will be interpreted as (1, 364), i.e. 52 weeks (the last day of
the last month will not be simulated). In Draft simulations, the simulation is always carried out on 8760 hours.

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

15/57

(Derated)
All time-series will be replaced by their mean and the number
of MC years is set to 1. If the TS are ready-made or Antaresgenerated but are not to be stored in the INPUT folder, the
mean time-series will not be written over the original ones. If
the time-series are built by Antares and if it is specified that
they should be stored in the INPUT, the single mean-time
series will be stored instead of the whole set of time-series.
Selection mode:
(Automatic)
All prepared MC years will actually be simulated.
(Custom)
The years to simulate are defined in a list. To set up this list,
choose in the main menu: Configure/ MC scenario playlist6.
Output profile
Simulation synthesis:

(True) Synthetic results will be stored in a directory :
Study_name/OUTPUT/simu_tag/Economy /mc-all

(False) No general synthesis will be printed out
Year-by-Year: (False) No individual results will be printed out
(True) For each simulated year, detailed results will be
printed out in an individual directory7 :
Study_name/OUTPUT/simu_tag/Economy /mc-i-number

Results Filtering:

(None) Storage of results for all areas, geographic districts,
interconnections as well as all time spans (hourly, daily, etc.)
(Custom) Storage of the results selected through “filters on
simulation results” of the Configure option in the main menu
Filters on areas, interconnections and time spans may also be
defined as follows:
a) On the map, select area(s) and/or interconnection(s)
b) Open the inspector module (Main menu, Windows)
c) Set adequate parameters in the “output print status” group

MC Scenarios:

(False) No storage of the time-series numbers (either
randomly drawn or user-defined) used to set up the simulation
(True) A specific OUTPUT folder will be created to store the
time-series numbers drawn when preparing the MC years

6
7

changing the number of MC years will reset the playlist to its default value ; not available in Draft simulations
Not available in Draft simulations
Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

16/57

RIGHT PART : Time-series management

For the different kinds of time-series that Antares manages in a non-deterministic way (load, thermal
generation, hydro power, wind power, solar power):
1) Choice of the kind of time-series to use
Either « ready-made » or «stochastic » (i.e. Antares-generated), defined by setting the status
parameter to “on” or “off”
2) For stochastic TS only:
Number
Number of TS to generate
Refresh
(Yes /No)
Indicates whether a periodic renewal of TS should be performed or not
Refresh span
Number of MC years at the end of which the renewal will be performed (if so required)
Seasonal correlation
(“monthly” or “annual”)
Indicates whether the spatial correlation matrices to use are defined month by month
or if a single annual matrix for the whole year should rather be used (see Section 6)
Store in input
(Yes/No)
Yes: the generated time-series will be stored in the INPUT in replacement of the
original ones (wherever they may come from)
No: the original time-series will be kept as they were
Store in output
(Yes/No)
Yes: the generated times-series will be stored as part of the simulation results
No: no storage of the generated time-series in the results directories
3) General rules for building up the MC years
Intra-modal

(Yes)
For each mode, the same number should be used for all locations (or 1 where there is
only one TS), but this number may differ from one mode to another. For instance,
solar power TS = 12 for all areas, while wind power TS number = 7 for all areas.
(No)

Inter-modal

Independent draws

(Yes)
For all modes, the same number should be used but may depend on the location (for
instance, solar and wind power TS = 3 for area 1, 8 for area 2, 4 for area 3, etc.)
(No)
Independent draws

A full meteorological correlation (for each MC year, one single number for all modes and areas) is,
from a theoretical standpoint, accessible by activating “intramodal” and “ inter-modal” for all but the
“thermal” kind of time-series. The availability of an underlying comprehensive multi-dimensional
Meteorological data base of ready-made time-series is the crux of the matter when it comes to using
this configuration.

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

17/57

User’s Notes

A built-in notepad for recording comments regarding the study. Such comments typically help
to track successive input data updates (upgrading such interconnection, removing such plant,
etc.). Another simple use is to register what has been stored in the “user” subfolder and why.
Such notes may prove useful to sort and interpret the results of multiple simulations carried
out at different times on various configurations of the power system.

Load
This window is used to handle all input data regarding load. In Antares load should include
transmission losses. It should preferably not include the power absorbed by pumped storage
power plants. If it does, the user should neither use the “PSP” array (see window “Misc. Gen”)
nor the explicit modeling of PSP plants
The user may pick any area appearing in the list and is then given access to different tabs :
•

The “time-series” tab display the “ready-made” 8760-hour time-series available for
simulation purposes. These data may come from any origin outside Antares, or be
data formerly generated by the Antares time-series stochastic generator, stored as
input data on the user’s request. Different ways to update data are :
o

direct typing

o

copy/paste a selected field to/from the clipboard

o

load/save all the time-series from/to a file (usually located in the “user”
subfolder)

o

Apply different functions (+,-, *, /,etc.) to the existing (possibly filtered) values
(e.g. simulate a 2% growth rate by choosing “ multiply-all-by-1.02”)

o

Handle the whole (unfiltered) existing dataset to either
• Change the number of columns (function name : resize)
• Adjust the values associated with the current first day of the year
(function name : shift rows)

Versatile “Filter” functions allow quick access to user-specified sections of data (e.g.
display only the load expected in the Wednesdays of January, at 09:00, for time-series
#12 to #19). Hourly load is expressed in round numbers and in MW. If a smaller unit
has to be used, the user should define accordingly ALL the data of the study (size of
thermal plants, interconnection capacities, etc.)
Note that:
If the “intra-modal correlated draws” option has not been selected in
the simulation window, MC adequacy or economy simulations can
take place even if the number of time-series is not the same in all
areas (e.g. 2 , 5 , 1 , 45 ,...)
If the “intra-modal correlated draws” option has been selected in the
simulation window, every area should have either one single timeseries or the same given number (e.g. 25 , 25 , 1 , 25...)

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

18/57

•

The “spatial correlation” tab gives access to the inter-area correlation matrices that will
be used by the stochastic generator if it is activated. Different sub-tabs are available
for the definition of 12 monthly correlation matrices and of an overall annual
correlation matrix.
A matrix A must meet three conditions to be a valid correlation matrix:
for all i and j { Aii= 100 , -100<= Aij <=100 } ; A symmetric ; A positive semi-definite

When given invalid matrices, the TS generator emits an unfeasibility diagnosis
•

The “local data” tab is used to set the parameters of the stochastic generator. These
parameters are presented in four sub-tabs whose content is presented in Section 6.

•

The “digest” tab displays for all areas a short account of the local data

Thermal
This window is used to handle all input data regarding thermal dispatchable power.
The user may pick any area appearing in the area list and is then given access to the list of
thermal plants clusters defined for the area (e.g. “CCG 300 MW”, “coal 600”,...). Once a given
cluster has been selected, a choice can be made between different tabs:
•

The “time-series” tab displays the “ready-made” 8760-hour time-series available for
simulation purposes. These data may come from any origin outside Antares, or be
data formerly generated by the Antares time-series stochastic generator, stored as
input data on the user’s request. Different ways to update data are :
o
o
o
o
o

direct typing
copy/paste a selected field to/from the clipboard
load/save all the time-series from/to a file (usually located in the “user”
subfolder)
Apply different functions (+,-, *, /,etc.) to the existing (possibly filtered) values
(e.g. simulate a 2% growth rate by choosing “ multiply-all-by-1.02”)
Handle the whole (unfiltered) existing dataset to either
• Change the number of columns (function name : resize)
• Adjust the values associated with the current first day of the year
(function name : shift rows)

Versatile “Filter” functions allow quick access to user-specified sections of data (e.g.
display only the generation expected on Sundays at midnight, for all time-series).
Hourly thermal generation is expressed in round numbers and in MW. If a smaller unit
has to be used, the user should define accordingly ALL the data of the study (Wind
generation, interconnection capacities, load, hydro generation, solar, etc.)

Note that:
If the “intra-modal correlated draws” option has not been selected in
the simulation window, MC adequacy or economy simulations can
take place even if the number of time-series is not the same in all
areas (e.g. 2, 5, 1, 45,etc.)
If the “intra-modal correlated draws” option has been selected in the
simulation window, every area should have either one single timeseries or the same given number (e.g. 25, 25, 1, 25, etc.). Note that,
unlike the other time-series (load, hydro, etc.), which depend on
meteorological conditions and are therefore inter-area-correlated, the
thermal plants time-series should usually be considered as
uncorrelated. Using the “correlated draws” feature makes sense only
in the event of having to play predefined scenarios (outside regular
MC scope)
Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

19/57

•

The “TS generator” tab is used to set the parameters of the stochastic generator.
These parameters are defined at the daily scale and are namely, for each day : the
average duration of forced outages (beginning on that day), the forced outage rate,
the duration of planned outages (beginning on that day), the planned outage rate,
planned outages minimum and maximum numbers. Durations are expressed in days
and rates belong to [0 , 1].

•

The “Common” tab is used to define the cluster’s techno-economic characteristics :
Name
Fuel used
Location (Area)
Activity status
•
false: not yet commissioned,moth-balled, ...
•
true : the plant may generate
Number of units
Nominal capacity
Full Must-run status
• false: above a partial “must-run level” (that may exist or not, see
infra) plants will be dispatched on the basis of their market bids.
•
true: plants will generate at their maximum capacity, regardless of
market conditions
Minimum stable power (MW)
Minimum Up time (hours)
Minimum Down time (hours)
Default contribution to the spinning reserve (% of nominal capacity)
CO2 tons emitted per electric MWh
Marginal operating cost (€/MWh)
Fixed cost (No-Load heat cost) (€ / hour of operation )
Start-up cost (€/start-up)
Market bid (€/MWh)
Random spread on the market bid (€/MWh)
Seasonal marginal cost variations (gas more expensive in winter, ...)
Seasonal market bid modulations (assets costs charging strategy )
Nominal capacity modulations (seasonal thermodynamic efficiencies, special
over-generation allowances, etc). These modulations are taken into account
during the generation of available power time-series
Minimal generation commitment (partial must-run level) set for the cluster
Note that:
The optimal dispatch plan as well as locational marginal prices are based
on market bids, while the assessment of the operating costs associated
with this optimum are based on cost parameters. (In standard “perfect”
market modeling, there is no difference of approaches because market bids
are equal to marginal costs)

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

20/57

Hydro
This window is used to handle all input data regarding hydro power
The user may pick any area appearing in the list and is then given access to different tabs:
•

The “time-series” tab displays the “ready-made” time-series already available for
simulation purposes. There are two categories of time-series (displayed in two
different subtabs): the Run Of River (ROR) time-series on the one hand and the
Storage power (SP) time-series on the other hand.

ROR time-series are defined at the hourly scale; each of the 8760 values represents
the ROR power expected at a given hour, expressed in round number and in MW. The
SP time-series are defined at the monthly scale; each of the 12 values represents the
overall SP energy expected in the month, expressed in round number and in MWh.
These data may come from any origin outside Antares, or be data formerly generated
by the Antares time-series stochastic generator, stored as input data on the user’s
request. Different ways to update data are:
o
o
o
o
o

direct typing
copy/paste a selected field to/from the clipboard
load/save all the time-series from/to a file (usually located in the “user”
subfolder)
Apply different functions (+,-, *, /,etc.) to the existing (possibly filtered) values
(e.g. simulate a 2% growth rate by choosing “ multiply-all-by-1.02”)
Handle the whole (unfiltered) existing dataset to either
• Change the number of columns (function name : resize)
• Adjust the values associated with the current first day of the year
(function name : shift rows)

Note that:
For a given area, the number of ROR time-series and SP timesseries must be identical

If the “intra-modal correlated draws” option has not been selected in
the simulation window, MC adequacy or economy simulations can
take place even if the number of hydro time-series is not the same in
all areas
(e.g.
2 , 5 , 1 , 45 ,...)
If the “intra-modal correlated draws” option has been selected in the
simulation window, every area should have either one single timeseries or the same given number (e.g. 25 , 25 , 1 , 25...)
•

The “spatial correlation” tab gives access to an annual inter-area correlation matrix
that will be used by the stochastic generator if it is activated. Correlations are
expressed in percentages, hence to be valid this matrix must be symmetric, p.s.d,
with a main diagonal of 100s and all terms lying between (-100 ,+100)

•

The “Allocation” tab gives access to an annual inter-area allocation matrix A(i,j) that is
used during the optimization process, regardless of whether the stochastic time-series
generator is used or not. This matrix describes the weights that are given to the loads
of areas (i) in the definition of the monthly and weekly hydro storage generation
profiles of areas (j).

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

21/57

More precisely, if there are Z zones z = 1,…,Z, and if M(z) denotes the overall “must-run”
generation of every kind in zone z (thermal, wind power, solar, run of river,…), one can
define successively :
a) L(z), “natural” load in zone z
b) L*(z) = L(z) - M(z), “net” load in zone z (in that sense that it has to be satisfied by
either imports or dispatchable thermal generation or hydro storage power).

c) L**(z) = sigma {i=1, Z} A(i,z) x L*(i) , “weighed” load in zone z. This weighed load
is the signal used in the hydro storage monthly and weekly energy profiles
adjustment stage, along with other parameters described further below.
Extreme cases are :
A is the identity matrix
The hydro storage energy monthly and weekly profiles of each zone z depend
only on the local demand and must-run generation in z
A has a main diagonal of zeroes
The hydro storage energy monthly and weekly profiles of each zone z do not
depend at all on the local demand and must-run generation in z

•

The “local data” tab is used to set up the parameters of the stochastic generator AND
to define techno-economic characteristics of the hydro system that are used in
Economy and Adequacy optimizations.
The parameters of the stochastic generator are the expectations,
standard deviations, minimum and maximum values of monthly energies
(expressed in GWh), monthly shares of Run of River within the overall
hydro monthly credit, and correlation between the energy of a month and
that of the next month (inter-monthly correlation).

These monthly energies will be considered either as amounts of energy to
generate or as amounts of energy inflows that may be partly stored in a
reservoir (i.e. net storable hydro energy = overall energy - ROR share)

The techno-economic characteristics used in optimizations are namely :
o

A “reservoir management” parameter that can take two values
(No / Yes). In the first case, the monthly SP time-series are
considered as energies to generate, while in the second they are
inflows to manage at best.

o

The reservoir capacity (or size S), in GWh. This parameter is not
used if “reservoir management” is set to No. If “reservoir
management” is set to Yes , the capacity must be strictly positive

o

12 x 3 values for the monthly reservoir levels at the beginning of
the month : low, average, high (expressed in percentage of
reservoir size : Ll, La,Lh)

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

22/57

o

o

365 x 3 values for the daily maximum available hydro SP Power,
in MW, deemed to be consistent with the three assumptions
regarding the level of the reservoirs: low, average, high
(Pl,Pa,Ph)
An “inter-monthly generation breakdown” parameter α, which is
used to split the hydro energy allocated for the whole year (sum
of inflows) into monthly energies, depending on the conditions
encountered at medium term (load level, wind generation,
etc...).This parameter is heuristically fitted and allows to simulate
different hydro management strategies. It is used only if the
hydro-time series are considered as inflows (otherwise, the timeseries are assumed to be direct estimates of monthly generated
energies)

The heuristic used comprises three steps:
a) Assessment of 12 monthly hydro storage energy targets H(m)
defined for each zone as follows : if H denotes the annual sum of
the 12 storable parts of the monthly inflows, L**(m) is the weighed
load during month m, then for all m and n :
H(m)/H(n) =( L**(m) / L**(n)) ^ α
Sigma {m=1,12} H(m) = H
b) Assessment of 12 monthly hydro storage energies H**(m) that
can actually be generated, by solving a linear problem in which
the objective function to minimize is a cost proportional to the
absolute deviations to the targets identified in step (a), while the
energies generated are variables submitted to two kinds of
constraints :
i. Energy conservation (monthly) : for any month ,
inflow - generation = monthly reservoir levels
variation
ii. Energy conservation (annual) : the energy
generated throughout the year is equal to the
sum of inflows8
iii. Reservoir constraints (soft) : the “low” and “high”
levels boundaries should not be crossed, or only
at a very high cost
iv. Reservoir constraints (hard) : reservoir level
always lies between (0, reservoir size)
c) Assessment of 365 daily maximum hydro-storage power levels
that will actually be used in the course of the simulations. These
power levels are the result of an interpolation between the low,
average and high arrays (Pl,Pa,Ph) made on the basis of the
reservoir levels that are secondary results of step (b)

8

This is equivalent to stating that the reservoir level at the end of the year is the same as at the beginning.
In the simulations, the initial reservoir level is randomly drawn between “L(low)” and “L(high)”, with average “L(average)”

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

23/57

o

An “inter-daily generation breakdown” parameter β, which is used
to split the hydro storage energy H**(m) to generate during the
whole month into daily energies, depending on the conditions
encountered at short term (load level, wind generation, etc...).This
parameter is heuristically fitted and allows to simulate different
hydro management strategies

The heuristic used comprises two steps :
a) Assessment of daily hydro storage energy targets h(i) defined for
each zone as follows: if H**(m) is the hydro storage energy to
generate during month m, l**(i) is the weighed load during day i,
then for all i and j :
h(i) / h(j) =( l**(i) / l**(j)) ^ β
Sigma {i=1,28>31} h(i) = H**(m)

b) Assessment of daily hydro storage energies h**(i) that can
actually be generated, by solving a linear problem in which the
objective function to minimize is a cost proportional to the
absolute deviations to the targets identified in step (a), while the
energies generated are variables submitted to two kinds of
constraints :
i. Energy conservation (monthly, soft) : The energy
generated throughout the month should be equal
to the sum of the daily targets. If it is lower (some
energy is spilled), the spillage cost translates as
a high penalty in the objective function
ii. Power constraints (hard) : for each day , the
energy generated lies between (0, P max x 24)

o

An “intra-daily modulation” parameter, which represents, for the
storage power, the maximum authorized value for the ratio of the
daily peak to the mean power generated throughout the day. This
parameter is heuristically fitted and allows to simulate different
hydro management strategies.
Extreme values are :
1:

Hydro storage power is constant throughout the day

24:

If the maximum hydro power is high enough, the
whole daily energy credit can be spent in one hour

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

24/57

Wind
This window is used to handle all input data regarding Wind power
The user may pick any area appearing in the list and is then given access to different tabs:
•

The “time-series” tab display the “ready-made” 8760-hour time-series already
available for simulation purposes. These data may come from any origin outside
Antares, or be data formerly generated by the Antares time-series stochastic
generator, stored as input data on user’s request. Different ways to update data are :
o
o
o
o
o

direct typing
copy/paste a selected field to/from the clipboard
load/save all the time-series from/to a file (usually located in the “user”
subfolder)
Apply different functions (+,-, *, /,etc.) to the existing (possibly filtered) values
(e.g. simulate a 2% growth rate by choosing “ multiply-all-by-1.02”)
Handle the whole (unfiltered) existing dataset to either
• Change the number of columns (function name : resize)
• Adjust the values associated with the current first day of the year
(function name : shift rows)

Versatile “Filter” functions allow quick access to user-specified sections of data (e.g.
display only the wind generation expected between 17:00 and 21:00 in February, for
time-series 1 to 100).
Hourly wind generation is expressed in round numbers and in MW. If a smaller unit
has to be used, the user should define accordingly ALL the data of the study (size of
thermal plants, interconnection capacities, load, etc.)
Note that:
If the “intra-modal correlated draws” option has not been selected in
the simulation window, MC adequacy or economy simulations can
take place even if the number of time-series is not the same in all
areas (e.g. 2 , 5 , 1 , 45 ,...)
If the “intra-modal correlated draws” option has been selected in the
simulation window, every area should have either one single timeseries or the same given number (e.g. 25 , 25 , 1 , 25...)
•

The “spatial correlation” tab gives access to the inter-area correlation matrices that will
be used by the stochastic generator if it is activated. Different sub-tabs are available
for the definition of 12 monthly correlation matrices and an overall annual correlation
matrix.
A matrix A must meet three conditions to be a valid correlation matrix:
for all i and j { Aii= 100 , -100<= Aij <=100 } ; A symmetric ; A positive semi-definite

When given invalid matrices, the TS generator emits an unfeasibility diagnosis
•

The “local data” tab is used to set the parameters of the stochastic generator. These
parameters are presented in four subtabs whose content is presented in Section 6.

•

The “digest” tab displays for all areas a short account of the local data

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

25/57

Solar
This window is used to handle all input data regarding Solar power. Both thermal generation
and PV generation are assumed to be bundled in this data section.

The user may pick any area appearing in the list and is then given access to different tabs :
•

The “time-series” tab display the “ready-made” 8760-hour time-series available for
simulation purposes. These data may come from any origin outside Antares, or be
data formerly generated by the Antares time-series stochastic generator, stored as
input data on the user’s request. Different ways to update data are :
o
o
o
o
o

direct typing
copy/paste a selected field to/from the clipboard
load/save all the time-series from/to a file (usually located in the “user”
subfolder)
Apply different functions (+,-, *, /,etc.) to the existing (possibly filtered) values
(e.g. simulate a 2% growth rate by choosing “ multiply-all-by-1.02”)
Handle the whole (unfiltered) existing dataset to either
• Change the number of columns (function name : resize)
• Adjust the values associated with the current first day of the year
(function name : shift rows)

Versatile “Filter” functions allow quick access to user-specified sections of data (e.g.
display only the solar power expected in August at noon, for all time-series).
Hourly solar power is expressed in round numbers and in MW. If a smaller unit has to
be used, the user should define accordingly ALL the data of the study (size of thermal
plants, interconnection capacities, etc.)
Note that:
If the “intra-modal correlated draws” option has not been selected in
the simulation window, MC adequacy or economy simulations can
take place even if the number of time-series is not the same in all
areas (e.g. 2 , 5 , 1 , 45 ,...)
If the “intra-modal correlated draws” option has been selected in the
simulation window, every area should have either one single timeseries or the same given number (e.g. 25 , 25 , 1 , 25...)
•

The “spatial correlation” tab gives access to the inter-area correlation matrices that will
be used by the stochastic generator if it is activated. Different sub-tabs are available
for the definition of 12 monthly correlation matrices and of an overall annual
correlation matrix.
A matrix A must meet three conditions to be a valid correlation matrix:
for all i and j { Aii= 100 , -100<= Aij <=100 } ; A symmetric ; A positive semi-definite

When given invalid matrices, the TS generator emits an unfeasibility diagnosis
•

The “local data” tab is used to set the parameters of the stochastic generator. These
parameters are presented in four subtabs whose content is presented in Section 6.

•

The “digest” tab displays for all areas a short account of the local data

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

26/57

Misc. Gen.
This window is used to handle all input data regarding miscellaneous non dispatchable
generation.
On picking any area in the primary list, the user gets direct access to all data regarding the
area, which amount to 8 ready-made 8760-hour time-series (expressed in MW) :
•

CHP generation

•

Bio Mass generation

•

Bio-gas generation

•

Waste generation

•

Geothermal generation

•

Any other kind of non dispatchable generation

•

A predefined time-series for the operation of Pumped Storage Power plants, if they
are not explicitly modeled. A positive value is considered as an output (generating) to
the grid, a negative value is an input (pumping) to the station.
Note that the sum of the 8760 values must be negative, since the pumping to
generating efficiency is lower than 1. The user may also use only the negative values
(prescribed pumping), while transferring at the same time the matching generating
credit on the regular hydro storage energy credit.

•

ROW balance: the balance with the rest of the world. A negative value is an export to
ROW, a positive value is an import from ROW. These values acts as boundary
conditions for the model

Different ways to update data are:
o
o
o
o
o

direct typing
copy/paste a selected field to/from the clipboard
load/save all the time-series from/to a file (usually located in the “user”
subfolder)
Apply different functions (+,-, *, /,etc.) to the existing (possibly filtered) values
(e.g. simulate a 2% growth rate by choosing “ multiply-all-by-1.02”)
Handle the whole (unfiltered) existing dataset to either
• Change the number of columns (function name : resize)
• Adjust the values associated with the current first day of the year
(function name : shift rows)

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

27/57

Reserves / DSM
This window is used to handle all input data regarding reserves and the potential of “smart”
load management (when not modeled using “fake” thermal dispatchable plants). On picking
any area in the primary list, the user gets direct access to all data regarding the area, which
amount to four ready-made 8760-hour time-series (expressed in MW). The first two are used
only in “draft” simulations, while the last two are available in either “adequacy” or “economy”
simulations:
•

Primary reserve: must be provided whatever the circumstances, even at the price of
some unsupplied energy (Draft simulations only)

•

Strategic reserve: sets a limit on the backup power that an area is supposed to be
able to export to its neighbours. This reserve may represent an actual generation
reserve, an energy constraint too complex to model by standard means (e.g. energy
policy regarding special reservoirs) or can also be justified by simplifications made in
grid modeling. (Draft simulations only).

•

Day-ahead reserve: power accounted for in setting up the optimal unit-commitment
and schedule of the following day(s), which must consider possible forecasting errors
or last-minute incidents. If the optimization range is of one day, the reserve will be
actually seen as “day-ahead”. If the optimization range is of one week, the need for
reserve will be interpreted as “week-ahead”. (Adequacy and Economy simulations)

•

DSM: power (decrease or increase) to add to the load. A negative value is a load
decrease, a positive value is a load increase. Note that an efficient demand side
management scheme may result in a negative overall sum (All simulation modes).

Links
This window is used to handle all input data regarding the interconnections. On picking any
interconnection in the primary list, the user gets direct access to all data regarding the link,
which are two annual parameters and a set of five ready-made 8760-hour time-series
The two parameters, used in economy or adequacy simulations (not in draft), are namely:
•

“ Hurdle cost ”, which is used to state whether (linear) transmission fees
should be taken into account or not in economy and adequacy simulations

•

“ Transmission capacities ”, which is used to state whether the capacities to
consider are those indicated in 8760-hour arrays or if zero or infinite values
should be used instead (actual values / set to zero / set to infinite)

The five times-series are:
•

NTC direct : the upstream-to-downstream capacity, in MW

•

NTC indirect : the downstream-to-upstream capacity, in MW

•

Impedances: virtual impedances that are used in economy simulations to give a
physical meaning to raw outputs, when no binding constraints have been defined to
enforce Kirchhoff’s laws (see “Output” section, variable “Flow quad”).

•

Hurdle cost direct : an upstream-to-downstream transmission fee, in €/MWh

•

Hurdle cost indirect : a downstream-to-upstream transmission fee, in €/MWh
Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

28/57

Binding constraints
This window is used to input all data regarding linear constraints that may have to be taken
into account in optimizations tasks.
In standard simulations, power is assumed to be able to flow freely within the grid without any
kind of constraints, aside from limits resulting from the interconnections capacities. This
modeling of the grid’s behavior allows the assessment of a rough adequacy/economic
diagnosis, in which all the theoretical potential of both grid and park are used to the end (say
that ad hoc FACTS come into action whenever needed to “abolish” Kirchhoff’s’ laws)
In more refined simulations, where the power system is simulated on an everyday “regular”
basis, the previous “emergency” modeling may be fitting no longer and, if enough data are
available, can be improved. For that purpose, Antares gives the possibility to define linear
constraints that are to be met by power flows while seeking the optimal solution of the
economic problem.
These constraints are freely defined by the user and can be arbitrarily mixed.
are available for the
For instance, if meaningful estimates of physical impedance
interconnections , Kirchhoff’s laws can be enforced by defining, for each element of a set of
cycles forming a basis of the system’s graph, the binding constraint:

∈

=0

Constraints serving different objectives may be defined, such as enforcing commercialcontract-related flow patterns.
Antares allows to define three categories of binding constraints:
•

“hourly” binding constraints, which are applied to instant power and which are typically
used for Kirchhoff’s laws enforcement

•

“daily” binding constraints, that are applied to daily energies. This class makes more
sense for commercial modeling (say : imports and exports from/to such and such
area should be comprised between such and such lower bound and upper bound).
Daily binding constraints may also be used to model more complex configurations,
such as setting up a virtual pumped storage power facility operated on a daily cycle

•

“weekly” binding constraints, that are applied to weekly energies. Like the previous
ones, these constraints may be used to model commercial contracts or various
phenomena, such as the operation of a pumped storage power plant operated on a
weekly cycle.

The main window is divided into six tabs whose purpose is:
•
•
•
•
•

•

“summary” : creates, edits or deletes a binding constraint of a given name (e.g.
“northern mesh”) and numerical class ( upper-bounded , lower-bounded, equality)
“weights” : defines how much each interconnection contributes to the constraint
(e.g. in the constraint 3 X(t) + 5.2 Y(t+4 ) = RHS, weights are 3 and 5.2)
“offsets” : defines whether the time-dependent variable X involved in the constraint is
related to time t or to another reference
(e.g. in the constraint 3 X(t) + 5.2 Y(t+4) =RHS, offsets are 0 and 4)
“=” defines the 8760-hour or 365- day right-hand side of the equality constraints
“>”defines the 8760-hour or 365-day right-hand side of the lower-bounded constraints
“<”defines the 8760-hour or 365-day right-hand side of the upper-bounded constraints

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

29/57

Note that the terms of the right-hand sides of the daily and weekly constraints are both defined
at the daily scale. For each simulation week, the terms
,
= 1,7 are interpreted as
follows:
a) If the type of the constraint is “daily” each
daily constraint

is the right-hand side value for the

b) If the type of constraint is “weekly” , the value used for week

is ∑

∈

When defining binding constraints between (hourly) power, daily or weekly (energy) flows,
attention should be paid to potential conflicts between them, since it is possible to create
situations for which the load/generation equilibrium problem has no solution, e.g.
( T1+T2 > 5 ; T2 < -3 ; T3 > 0 ; T1+T3 < 7)

On running the economic simulator, such situations will produce an unfeasibility diagnosis.

Economic Opt.
This window is used to set the value of a number of area-related parameters that, aside from
the costs of each generating plant, define the optimal solution that Antares has to find in
economic simulations. These parameters are namely, for each area of the system :
•

The value of the unsupplied energy (also commonly denoted Value Of Lost Load,VOLL) , in
€/MWh. This value should usually be set much higher than the cost of the most expensive
generating plant of the area

•

The random spread within which the nominal unsupplied energy value is assumed to vary

•

The value of the spilled energy, in € /MWh. This value reflects the specific penalty that should
be added to the economic function for each wasted MWh, if any. Note that even if this value
is set to zero no energy will be shed needlessly

•

The random spread within which the nominal unsupplied energy value is assumed to vary

•

Three parameters named “shedding status” and related to different kinds of generation. If the
system cannot be balanced without shedding some generation, these parameters give control
on how each kind of generation ("Non dispatchable power", ”Dispatchable hydropower” and
“Other dispatchable generating plants”) should contribute to the shedding. Depending on the
value chosen for the status, the generation can or cannot be shed to find a solution to the
load/generation balance problem. Note that enforcing a negative status for all types of plants
may lead to simulations scenarios for which there are no mathematical solutions.
On running the economic simulator, such situations produce an unfeasibility diagnosis.

Miscellaneous
In all previous windows showing Input data, the content can be filtered so as to reflect only items that
are associated with Areas and Links defined as “visible” in a particular map. In that regard, binding
constraints are considered as visible if and only if all of their non-zero weight associated objects are
visible on the map.

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

30/57

5 Output files
The general file organization is the same for Economy, Adequacy and Draft simulations.
•
•

Economy and Adequacy results may be displayed in the GUI ( “Output” in main menu)
Draft results are available only as flat .txt files. They can be viewed with “Tool /csv viewer” in
the main menu (As well as any other files, they can also be accessed by Xcel or suchlike)

Economy :
OUTPUT/ Simu id / Economy /mc-all

/ grid /...
contains a summary file "digest.txt"
/areas/name/...
contains area-related results
/links / name/...
contains interconnection-related results
/mc-ind /
/areas/name/... contains area-related results
/links / name/... contains interconnection-related results

(“mc-all” files contain synthetic results over all years, “year-number” files contain results for a single year
The variables present in each file are detailed in the following sections
In “Economy” simulations, all variables have a techno-economic meaning

Adequacy:
OUTPUT/ Simu id / Adequacy /mc-all / grid /...
contains a summary file "digest.txt"
/areas/name/...
contains area-related results
/links / name/...
contains interconnection-related results
/mc-ind /
/areas/name/... contains area-related results
/links / name/... contains interconnection-related results
(“mc-all” files contain synthetic results over all years, “year-number” files contain results for a single year
The variables present in each file bear exactly the same name as in Economy simulations but do not have the same values
The only variables that have a techno-economic meaning are the “Adequacy” indicators (unsupplied energy,LOLD,LOLP)

Draft:

OUTPUT / Simu id / Adequacy-Draft / mc-all /grid/...
/areas/name/...

contains a condensed file "digest.txt"
contains area-related results

(“mc-all” files contains mostly synthetic results over all years ; However, there is (for each area) a “mc-annual.txt” file that gives
a short view of local results for each simulated year)

IMPORTANT Adequacy and Economy files look the same but their content are specific
In “Economy” and “Adequacy” simulations, the optimization ignores the “primary” and “strategic”
reserves (however, it may include the [other] spinning and day-ahead reserves, depending on the
settings made in “optimization preferences”).
In “Adequacy” simulations, all dispatchable thermal units are given the “must-run” status (hence, they
will generate at Pmax, regardless of the demand). As a consequence the only variables that are
actually meaningful are the adequacy indicators (unsupplied energy, LOLD,LOLP), that may depend
on assumptions made regarding the economic values of Unsupplied and spilled energies, and on
hurdle costs on interconnections.
As a consequence, both “Adequacy” and “Economy” simulations yield the same values for the
adequacy indicators under the following conditions: if hurdle costs on interconnections are higher than
the difference between the maximum VOLL and the minimum VOLL assigned to the different areas of
the system.
The files and their content are hereafter described.
Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

31/57

Economy and Adequacy, area results9
15 files resulting from the combination of the following attributes :
[values | id | details] X [hourly | daily | weekly | monthly | annual]
•
•

The second attribute defines the time span over which the results are assessed : hourly detail,
daily bundle, weekly bundle, monthly bundle, annual bundle.
The first attribute defines the nature of the results presented in the file :

Values

Values of different variables (price, load, overall generation issued from coal, etc.),
the list of which is common to all areas of the interconnected system. Files of type
"values" have therefore the same size for all areas.
These results appear under the label “ general values” in the output GUI

details

Values regarding the different dispatchable thermal generating plants of each area
(e.g. “older 300 MW coal from the south coast”). The sizes of these files differ from
one area to another.
These results appear under the label “ thermal plants ” in the output GUI

id

Identifier (number) of the Monte-Carlo years for which were observed the extreme
values of the different variables presented in the « values » files
These results appear under the label “record years” in the output GUI

The area files that belong to the « values » class display 98 fields corresponding to the expectation ,
standard deviation, minimal and maximal values of the variables whose list is given hereafter.
OV.COST
OP.COST
MRG. PRICE
CO2 EMIS.
BALANCE
ROW BAL
PSP
MISC. NDG
LOAD
H.ROR
WIND
SOLAR
NUCLEAR
LIGNITE
COAL
GAS
OIL
MIX.FUEL
MISC.DTG
H.STOR
UNSP. ENRG
SPIL. ENRG
LOLD
LOLP
AVL DTG
DTG MRG
MAX MRG
NP COST
NODU

Overall cost = operating cost + unsupplied energy cost+ spilled energy cost
Operating cost = Proportional costs + Non- proportional costs
LMP : overall economic effect of a local 1MW load increase
Amount of CO2 emitted by all dispatchable thermal plants
Overall Import/export balance of the area (positive value : export)
Import/export with areas outside the modeled system (positive value: import)10
User-defined settings for pumping and subsequent generating
Miscellaneous non dispatchable generation
Demand (including DSM potential if relevant)
Hydro generation, Run-of-river share
Wind generation
Solar generation (thermal and PV)
Overall generation of nuclear clusters
Overall generation of dispatchable thermal clusters burning brown coal
Overall generation of dispatchable thermal clusters burning hard coal
Overall generation of dispatchable thermal clusters burning gas
Overall generation of dispatchable thermal clusters using petroleum products
Overall gen. of disp. thermal clusters using a mix of the previous fuels
Overall gen. of disp. thermal clusters using other fuels
Hydro generation, storage share
Unsupplied energy : adequacy indicator (Expected Energy Not Served–EENS)
Spilled energy (energy that cannot be used and has to be wasted)
Loss of load duration :
adequacy indicator (length of shortfalls)
Loss of Load probability :
adequacy indicator (probability of shortfalls)
Available dispatchable thermal generation (sum of av. power over all plants)
Disp. Ther. Gen. (AVL DTG – sum of all dispatched thermal generation)
Maximum margin : operational margin obtained if the hydro storage energy of
the week were used to maximise margins instead of minimizing costs
Non-proportional costs of the dispatchable plants (start-up and fixed costs)
Number Of Dispatched Units11

9

This description applies to both « MC synthesis » files and “Year-by-Year” files, with some simplifications in the latter case
Value identical to that defined under the same name in the “Misc Gen” input section.
11
NODU and NP Cost do not appear in “Adequacy” results since these variables are irrelevant in that context
10

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

32/57

Economy and Adequacy, interconnection results12
10 files resulting from the combination of the following attributes :
[values | id] X [hourly | daily | weekly | monthly | annual]
•
•

The second attribute defines the period of time over which the results are assessed : hourly
detail, daily bundle, weekly bundle, monthly bundle, annual bundle.
The first attribute defines the nature of the results presented in the file

values

values of different variables (flow, congestion rent) the list of which is common to all
interconnections. The files of type "values" have therefore the same size everywhere
These results appear under the label “ general values” in the output GUI

id

identifier (number) of the Monte-Carlo years for which were observed the extreme
values of the different variables presented in the « values » files
These results appear under the label “record years” in the output GUI

The area files that belong to the « values » class display 27 fields corresponding to the expectation ,
standard deviation, minimal and maximal values of the variables whose list is given hereafter.
FLOW LIN.
Flow (signed + from upstream to downstream) assessed by the linear optimization.
These flows follow Kirchhoff’s law only if these laws have been explicitly enforced by the means of
suitable binding constraints
UCAP

Used capacity: absolute value of FLOW LIN. This indicator may be of interest
to differentiate the behavior of interconnectors showing low average flows: in
some cases this may indicate that the line is little used, while in others this
may be the outcome of high symmetric flows

FLOW QUAD.

Flow computed anew, starting from the linear optimum, by minimizing a
quadratic function equivalent to an amount of Joule losses, while staying
within the transmission capacity limits. This calculation uses for this purpose
the impedances found in the “Links” Input data. If congestions occur on the
grid, these results are not equivalent to those of a DC load flow

CONG. FEE ALG

Algebraic congestion rent = linear flow * (downstream price – upstream price)

CONG. FEE ABS

Absolute congestion rent = linear flow* abs(downstream price–upstream price)

MARG. COST

Decrease of the system’s overall cost that would be brought by the optimal
use of an additional 1 MW transmission capacity (in both directions)

CONG PROB +

Up>Dwn Congestion probability = (NC+) / (total number of MC years) with:
NC+ = number of years during which the interconnection was congested in the
Up>Dwn way for any length of time within the time frame relevant with the file

CONG PROB -

Dwn>Up Congestion probability = (NC-) / (total number of MC years) with:
NC- = number of years during which the interconnection was congested in the
Dwn>Up way for any length of time within the time frame relevant with the file

HURD. COST

Contribution of the flows to the overall economic function through the “hurdles
costs” component. For each hour :
if FLOW.LIN > 0
HURD. COST = (hourly direct hurdle cost)
*
(FLOW LIN.)
else
HURD.COST = (hourly indirect hurdle cost)
* (-1)* (FLOW LIN.)

12

This description applies to both « MC synthesis » files and “Year-by-Year” files, with some simplifications in the latter case

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

33/57

Economy and Adequacy, other results
Depending on the options chosen in the main simulation window, the output folders may
include either, both or none of the following sections :
•

OUTPUT/ Simu Id / ts-numbers / Load
/area names / ...
/ Thermal
/area names / ...
/ Hydro
/area names / ...
/ Wind
/area names / ...
/ Solar
/area names / ...
These files contain, for each kind of time-series, the number drawn (randomly or not) in each
Monte-Carlo year (files are present if “output profile / MC scenarios” was set to “true”)

•

OUTPUT/ Simu Id / ts-generator / Load
/ Thermal
/ Hydro
/ Wind
/ Solar

/ batch number /area names / ...
/ batch number /area names / ...
/ batch number /area names / ...
/ batch number /area names / ...
/ batch number /area names / ...

These files contain, for each kind of Antares-generated time-series, copies of the whole set of
time-series generated. Batch numbers depend on the values set for the “refresh span”
parameters of the stochastic generators (files are present if “store in output” was set to “true”)

Draft, area results
1 file « annual » + 6 files resulting from the combination of the following attributes :
[with-network | without-network | id] X [ hourly | annual]
•

The second attribute defines the period of time over which the results are assessed : hourly
detail or annual summary.

•

The first attribute defines the nature of the results presented in the file

with network

values of adequacy indices (shortfall duration, loss of load probability)
assessed while taking into account the effective grid capacities. The results in
these files bear the suffix –CN (connex)

without network

values of adequacy indices (shortfall duration, loss of load probability)
assessed without taking into account any interconnection. The results in these
files bear the suffix –IS (isolated areas)

id

identifiers (numbers) of the MC years for which were observed the extreme
values of the different variables presented in the « w/net » and “wo/net” files

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

34/57

Files « with network » et « without network » present the expectations and extreme values observed
for the variables whose list is given hereafter :
LOLD

Overall length of time for which there were shortfalls (Loss of Load Duration)
(note : the commonly used LOLE index is equivalent to LOLD expectation )

LOLP

Loss of Load Probability

EENS

Energy Not Supplied

MARG

Margin = available generation – (load + primary reserve)
When MARG>0, MARG is a security margin
When MARG <0 , MARG is a curtailment depth

The file « annual » has one line per simulated Monte-Carlo year and gives, for each year, the following
information :
LOLD IS

Load shedding duration, if the grid capacities are not considered as available

LOLD CN

Load shedding duration, if the grid capacities are actually available

MAX DEPTH IS

Margin available at the most critical hour of the whole MC year, w/o grid
When MAX DEPTH >0 , MAX DEPTH is a security margin
When MAX DEPTH <0,MAX DEPTH is a shortfall depth

MAX DEPTH CN

Margin available at the most critical hour of the whole MC year, w/ grid
When MAX DEPTH >0 , MAX DEPTH is a security margin
When MAX DEPTH <0,MAX DEPTH is a shortfall depth

Remark : In spite of their likenesses, the fields « MARG » of the files w/net, wo/net and the fields
« MAX DEPTH » of the file mc-details are not identical (hence different names):
•
•

MARG (expectation, min, max) is related to the whole set of MC years
MAX DEPTH regards one single year.

Note that the following relations hold :
Min { MC years } MAX DEPTH IS
Min { MC years } MAX DEPTH CN

= Min { hours}
= Min { hours}

MARG IS
MARG CN

[MIN]
[MIN]

Miscellaneous
Alike Input data, output results can be filtered so as to include only items that are associated with
Areas and Links defined as “visible” in the current map. In addition, the output filtering dialog box
makes it possible to filter according to two special categories (Districts and Unknown) that are not
related to standard maps:
•

•

Districts displays only results obtained for spatial aggregates
Unknown displays only results attached to Areas or Links that no longer
exist in the Input dataset (i.e. study has changed since the last simulation)

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

35/57

6 Time-series analysis and generation
General
When ready-made time-series are not available or are too scarce for building the required number of
Monte-Carlo annual simulation scenarios, Antares provides means to generate sets of stochastic timeseries to use instead.
The different categories of time-series call for wholly different generation processes :
•

For thermal power, the generator is based on the animation of a daily three-state Markov
chain (available – planned outage – forced outage ) attached to each plant.

•

For Hydro-power, the generator works out monthly time-series of energies, based on the
assumption that they can be modeled by Log Normal variables with known correlations
through space and time. So as to keep the model simple, for an interconnected system made
of N areas, the user defines, along with the N expectations and N standard deviations of the
monthly energies, the N X N correlation matrix R(n,m) of the logs of the annual hydro
energies between the areas n,m, and the N average auto-correlations r(k) between one
month and the next in each area k. The correlation C(n,i,m,j) between the logs of hydro
energies in area n, month i and area m, month j is taken to be C(n,i,m,j)=
R(n,m)*sqrt((r(n)*r(m))^abs(j-i)) This most simplified model asks for considerably fewer data
than a comprehensive 12N X 12N time-space matrix. Note that if R is positive semi-definite
but C is not, matrix C is automatically transformed into a fitting p.s.d matrix and the data
generation keeps going on (however, the log report will show a warning message). If the
primary matrix R is not p.s.d, data are considered as corrupted, the generation stops and a
fatal error message will be displayed in the log report

•

For Wind power, Solar power and Load, the required time-series are 8760-hour long and
have to emulate as closely as possible the response of the system to variations of wind
speed, sunshine and temperature. In all three cases, the rationale of the model is to offer the
possibility to consider either the final variable to model (wind power output, solar power
output, load) or an underlying intermediate variable (wind speed, nebulosity, deviation
between load and the level expected in standard temperature conditions) as a stationary
stochastic process, with given marginal laws, given auto-correlation functions and given
spatial correlations (eventually, the values of the final variables and those of the core
stationary process are tied by diurnal/seasonal rhythms and scaling functions).

The identification of all relevant parameters can be made outside Antares by any appropriate
means but can also be made automatically by the time-series analyzer, which is then to be fed
with the largest available set of historical time-series. Note however that, using the time-series
analyzer, one has to consider whether the time-series at hand are statistically meaningful or
whether they need some pre-processing (for instance, if wind power time-series are gathered
for a period within which the park has increased, the time-series to analyze should be
expressed in % of installed power rather than in MW. For Solar power, the relevant variable to
model as a stationary stochastic process is certainly not the raw output of solar power but
rather a meteorological indicator related to the sky clarity (for instance , time-series of
nebulosity expressed on a 0-100 scale may be used).
Once generated by appropriate algorithms, the values of the stationary processes are turned
into final values by using a number of parameters that put back in the series the diurnal and
seasonal patterns that may have been observed in the course of the historical data analysis
and that were temporarily removed to identify the core stationary processes.

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

36/57

Time-series generation (load, wind, solar) : principles

For the generation of wind, solar and load time-series, Antares gives access to different marginal laws
and autocorrelation functions presented hereafter. Note that wind speed modeling should usually be
based upon a Weibull modeling, while almost all other situations are likely to be best modeled by Beta
variables.
The stationary processes are defined at a monthly scale. For each month, there are :
•

Four parameters for the definition of the marginal law

Law
Uniform
Beta
Normal
Weibull
Gamma

Uniform :
Beta
:
Normal :
Weibull :
Gamma :

TS Gen. Parameters
α
β
γ
δ
N/A
N/A
<δ
>γ
>0
>0
<δ
>γ
Any
>0
N/A
N/A
>=1 <50
>0
N/A
N/A
>=1 <50
>0
N/A
N/A

Expectation

Variance

(δ−γ)/2

(δ−γ)^2 /12

γ + α(δ−γ)/(α+β)

[αβ(δ−γ)^2] / [(α+β+1)(α+β)^2 ]

α
β Γ(1+1/α)
α∗β

β^2
β^2[ Γ(1+2/α)−Γ(1+1/α)^2)]
α∗β^2

uniform defined on (γ,δ)
Beta(α,β) defined on (γ,δ)
expectation α, standard deviation β
shape α, scale β, defined on (0,+inf.)
shape α, scale β, defined on (0, +inf.)

In the expressions of expectation and variance, Γ(x) is the standard Euler Function
•

Two parameters for the definition of the autocorrelation function

Law
Pure exponential decay
Smoothed exponential
decay(*)

TS Gen. Parameters
θ
µ
θ>0
µ =1
θ>0
1 < µ < 24

Corr (Xt,Xt+h)
exp( - θh)
Phi(θ,µ,h)

Phi(θ,µ,h)= (1/A) *sigma {i=0,µ} [sigma {j=h,h+µ} (exp(-θ|j-i|))]
with A= µ + 2 sigma{i=1,µ ;j=1,µ ; j>i } (exp(-θ(j-i))
(*) Obtained by the generation of purely exponentially autocorrelated values (parameter θ) followed
by a moving average transformation (parameter µ). θ and µ should be carefully chosen so as to
accommodate at best the experimental data at hand. If meaningful historical data are available, this
identification may be directly made using the Antares time-series analyzer

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

37/57

Time-series generation (load, wind, solar) : GUI

The section of the GUI specific to the generation of wind, solar and load time-series comprises :
1. Spatial correlation matrices that are located within the “spatial correlation” tab
each path “ Wind | Solar | Load / ”

of

This tab contains a workspace for the description of 12 monthly spatial correlation matrices Ξ
and one annual correlation matrix. For the stochastic generators to work properly, these
matrices must meet the usual requirements (matrices must be p.s.d, symmetric, with all terms
between -100 and +100, and a main diagonal made of 100s). If this is not the case,
generators will emit an unfeasibility diagnosis. Matrices can be either set up manually OR
automatically filled out by the time-series analyzer (see next paragraph).

Depending on the choices made in the main “simulation” window, the matrices used will be
either the 12 monthly matrices or the annual matrix. Whether to use the first or the second
option depends on the quality of the statistical data at hand: with high quality data (for
instance, that derived from the analysis of a very large pool of historical data), use of monthly
correlations is recommended because monthly differences between matrices have a physical
meaning ; with less robust data (derived from a handful of historical data,…), use of the single
annual correlation matrix should be preferred because it smoothes out the numeric noise
which impairs the monthly matrices.

2.

Four parameters and four subtabs that are located within the “local” tab of each path
“Wind | Solar | Load / ”

FOUR PARAMETERS
•

Capacity : This first parameter is used to scale up time-series generated on the basis of the
(α,β,γ,δ,θ,µ) parameters described previously in the “principles” paragraph, together with
coefficients characterizing the diurnal pattern (see below)

•

Distribution : This second parameter gives the type of marginal distribution of the stationary
stochastic processes to generate (Beta, Weibull, Normal, Gamma, Uniform)

•

Translation : This third parameter has three possible values :
o
o

o

Do not use : parameter ignored
Add before scaling : A specific 8760-hour array is added to the time-series
by the primary stochastic generator, BEFORE use of the conversion table
followed by the final multiplication by the capacity factor
Add after scaling : A specific 8760-hour array is added to the time-series
by the primary stochastic generator, AFTER use of the conversion table
followed by the final multiplication by the capacity factor

produced
(optional)
produced
(optional)

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

38/57

•

Conversion : This fourth parameter has two possible values :
o

Do not use : Any transfer function that may be described in the “conversion” subtab
(see below) should not be used for the final stage of data elaboration (for instance, if
the primary parameters describe the physics of wind speeds, the time-series
eventually produced should remain wind speeds and not wind power).

o

Use : The time-series produced by the stochastic generators (wind speeds, for
instance) are turned into other values (wind power) by using the transfer function
described in the “conversion” subtab.

FOUR SUBTABS
•

Subtab “Coefficients”
A twelve-month table of values for the primary parameters α,β,γ,δ,θ,µ
This table may be either filled out manually or automatically (use of the time-series analyzer)

•

Subtab “Translation”
Contains a 8760-hour array T to add to the time-series generated, prior or after scaling. This
array can be either filled out manually or by the time-series analyzer

•

Subtab “Daily profile”
A 24*12 table of hourly / monthly coefficients K(hm) that are used to modulate the values of
the stationary stochastic process by which the actual process is approximated. This table can
be either filled out manually or by the time-series analyzer

•

Subtab “Conversion”
A table of 2 * N values (with 1<=N<=50 ) that is used to turn the initial time-series produced by
the generators (for instance, wind speeds) into final data (for instance , wind power). The
transfer function (speed to power,…) is approximated by N discrete points whose abscises
X(N) an ordinates Y(N) are given by the table.

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

39/57

Time-series analysis (load, wind, solar)

The time-series analyzer module available in Antares is meant to identify the values that should be
given to the parameters used in the time-series generators (load, solar power, wind power) so as to fit
best historical time-series at hand
IMPORTANT
When the time-series analyzer is used, it automatically updates the parameters relevant to the
analysis (for instance: analysis of “wind” time-series will overwrite all local and global “wind”
parameters [correlation matrices] that may have been previously set manually)
The primary TS analyzer window shows two tabs:
1. Tab “Time-series and areas”
o

Time-series (load, wind, solar) : class of parameters to be assessed by the analyzer

o

Browse: location of the historical time-series files. These are txt files in which 8760hour time-series must be stored in adjacent columns separated by a tabulation

o

For each area :
Activity status
• yes : parameters will be assessed and updated by the analyzer
• no : the area will be skipped (local parameters for the area will
remain unchanged, however spatial correlation with other areas will
be reset to zero)
Distribution
• Type of distribution to fit (beta, normal, etc.)
Data
•
•

Raw : data to analyze are the actual historical time-series
Detrended : data to analyze are the time-series of the deviations to
average (for instance: load time-series need to be analyzed in
“detrended” mode while wind speeds can be analyzed in “raw” mode)

File to analyze
• Name of the file that should contain historical time-series to analyze
Status
• Ready (a file bearing the expected name was found)
• Not found (no file found with the expected name)
IMPORTANT : To generate stochastic data similar to the historical data analyzed, generation
parameters must be kept consistent with the results of the analysis , which means, in the generators :
Keep the same:
Type of distribution
Values for α,β,γ,δ and for the diurnal–seasonal pattern (table of 12 X 24 values)
Value for the “capacity” parameter (the analyzer automatically sets it to 1)
Besides:

“Conversion” option must be set to “no”
“Translation” option must be set to “do not use “ if data were analyzed as “raw”
and to “add after scaling” or “add before scaling” if data were analyzed as
“detrended” (both options give the same value in this case because the scaling is 1:1)

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

40/57

2. Tab “Global settings”
o

Temporary folder : workspace that can be used for the analysis (cleaned after use)

o

Analyzer settings
Short- term autocorrelation adjustment (%)
Long – term autocorrelation adjustment (%)
These two parameters are used by Antares as targets for the fitting of θ and µ
parameters. For instance, if the historical time-series autocorrelation function is
such that Corr(T,T+ 18 hours)=90 % and Corr(T,T+60 hours)= 50%, and if the
parameters in the analyzer are (ST = 90%,LT = 50%) , then it will search values of
θ and µ matching the historical autocorr.function in two points(18 hours, 60 hours)
Trimming threshold (%)
In the spatial correlation matrices, terms lower than the threshold will be
replaced by zeroes

o

Input data
Time-series per area (n)
limits the analysis to the first n historical time-series at hand
Upper-bound (Max)
In the analysis, all values above Max in the historical files will be
replaced by Max
Lower-bound (Min)
In the analysis, all values below Min in the historical files will be
replaced by Min

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

41/57

Time-series generation (thermal)

The stochastic generator for time-series of available dispatchable power generation works, for
each plant of each set (cluster), with the following parameters :
•

The nominal plant capacity and a 8760-hour array of modulation
coefficients to apply to it (default value : 1)

•

A 365-day array of forced outages rates (“FOR” , lies in [0,1] )

•

A 365-day array of planned outages rates (“POR” , lies in [0,1])

•

A 365-day array of forced outages average durations
(“FOD” in days, integer, lies in [1,365])

•

A 365-day array of planned outages average durations
(“POD” in days, integer,lies in [1,365] )

•

A 365-day array of planned outages minimum number (PO Min Nb)
(integer, lies in [0, PO Max Nb])

•

A 365-day array of planned outages maximum number (PO Max Nb)
(integer, lies in [PO Min Nb, Nb of units in the cluster]

•

Two parameters describing how forced outages durations may
randomly deviate from their average value (law : uniform or geometric
, volatility : lie in [0,1])

•

Two parameters describing how planned outages durations may
randomly deviate from their average value (law : uniform or geometric
, volatility : lie in [0,1])

1. Outage duration : meaning and modeling
In the thermal time-series generator, the concept of outage duration (either forced or planned)
is simple enough : for any given plant affected by such an event, it is the duration of a single
outage, expressed in days.
The fact that 365 different values can be used to describe what may happen in the course of a
year (for each kind of outages) means that the average outage duration may depend on the
day the outage begins on. For instance, very short outages may be sometimes be planned on
week-ends. Likewise, historical statistics can show that forced outages do not last the same
average time in winter and summer, etc.
In complement to the average value of the duration D of outages beginning on a particular
day, the time-series generator allows to set two parameters that describe how the actual
outage durations may deviate from the calendar-related average value.

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

42/57

o

The first parameter (law) can take either the value “uniform” or “geometric” :
Uniform : the actual outage duration will be randomly drawn (one draw per
outage), according to a uniform distribution centered on the average value
D. The width of the interval [min duration, max duration] will depend on the
value of the second parameter (volatility)
Geometric : the actual outage duration will be expressed as the sum of a
fixed value F and a randomly drawn (one draw per outage) variable following
a geometric distribution of expectation G, with F+G=D. The ratio of F to G
will depend on the value of the second parameter (volatility).

o

The second parameter (volatility) can take any value within [0,1]
0:

The outage duration does not show any stochastic fluctuation at all.
Therefore, regardless of the chosen distribution law :
actual duration = D

1:

The variability of the actual outage duration is as high as the chosen
distribution law makes it possible, which means respectively that :
If choice = “uniform” : 1 <= actual duration <= 2D-1
If choice = “geometric” : F=0 and G = D
(which in turn implies 1<=actual duration <= #4D )

012

Minimum
1
1
1
1
1
1
1
1
1
1
1
1
1

Simulation Cores
Low
Medium
Large
1
1
1
1
1
2
2
2
2
2
2
3
2
3
4
2
3
4
2
3
5
2
4
6
3
5
7
3
5
8
3
6
8
3
6
9
Ceil(S/4)
Ceil(S/2)
Ceil(3S/4)
CPU allowances in parallel mode

Maximum
1
2
3
4
5
6
7
8
8
9
10
11
S-1

Note: The number of independent threads actually launched by Antares in parallel mode may
appear smaller than that shown in the table above. In this case, the resources monitor menu
and the dashboard displayed on starting the simulation indicates:
simulation cores: nn reduced to pp
nn is the regular allowance and pp is the practical value that the solver has to work with.
Allowance reduction may occur if the built-in Time-Series generators are activated, their
“refresh” status is set to “Yes” and the values given to the “refresh span” parameters are not
appropriate (parallel execution demand that refresh operations do not take place within a
bundle of parallel years). Optimal use of the “parallel” execution mode is obtained when all
activated built-in time –series generators are set up in either of the two following ways:
Refresh status : No
Refresh status : Yes, refresh span = Ki * (CPU allowance) , with Ki >=1
Examples of reduction from an initial allowance of 12 cores are given hereafter. The reduced
allowance is the size of the smallest bundle of parallel years between two consecutive
“refresh” (it indicates the slowest point of the simulation16). Note that RAM requirements
displayed in the resources monitor are, contrariwise, assessed on the basis on the largest
bundle of parallel years encountered in the simulation).
Built-in TS generators status / refresh span17
Load
Thermal
Hydro
Wind
Solar
50
1
50
50
50
No
10
50
No
No
No
11
50
No
No
No
100
100
100
100
No
12
12
12
No
12
24
48
48
36

Reduced Allowance (from 12)
MC Years : 80
MC years: 400
1
1
10
10
5
1
No reduction
418
No reduction
No reduction

15
This hardware characteristic, independent from Antares general parameters and from study parameters, can be checked with the Resources
monitor tool (Section 3)
16
When the number of MC years to run is smaller than the allowance, the parallel run includes all of these years in a single bundle and there
is no “reduced allowance” message
17
The Table indicates either the refresh status (No) or the refresh span (the associated refresh status “yes” is implicit)
18
The smallest bundle in this case is the ninth (year number 97 to year number 100). The first 8 bundles involve 12 MC years each

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

56/57

APPENDIX : ATTRIBUTION NOTICES
Antares_Simulator
Copyright 2007-2018 RTE - Authors: The Antares_Simulator Team
Antares_Simulator is free software: you can redistribute it and/or modify it under the terms of
the GNU General Public License as published by the Free Software Foundation, either version
3 of the License, or (at your option) any later version.
There are special exceptions to the terms and conditions of the license as they are applied to
this software. View the full text of the exceptions in file COPYING.txt in the directory of a
distribution of this software in source form.
Antares_Simulator is distributed in the hope that it will be useful, but WITHOUT ANY
WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A
PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with
Antares_Simulator. If not, see .
Antares_Simulator 6.0.0 uses external libraries and makes extensive use of the
following persons' or companies code. Source and binary forms of these programs are
distributed along with Antares_Simulator with NO WARRANTY:
Wxwidgets 3.0.2
Copyright (c) 1998-2017
The wxWidget Team
license: wxWindows Library License,V3.1
https://spdx.org/licenses/wxWindows.html
libCurl 7.51.0 Copyright (c) 1996-2017
license: curl license

Daniel Stenberg et al
https://spdx.org/licenses/curl.html

OpenSSL
1.1.0 Copyright (c) 1998-2016 The OpenSSL Project
"This product includes software developed by the OpenSSL Project
for use in the OpenSSL Toolkit (http://www.openssl.org/)"
"This product includes software written by Tim Hudson (tjh@cryptsoft.com)"
license: OpenSSL license and SSLeay license https://spdx.org/licenses/OpenSSL.html
libYuni 1.1.0
license:

https://github.com/libyuni
Mozilla Public License 2.0

https://spdx.org/licenses/MPL-2.0.html

Mersenne Twister
Copyright (c) 1997-2002 M.Matsumoto and T.Nishimura
license: 3-clause BSD
https://spdx.org/licenses/BSD-3-Clause.html
strtod library

Copyright (c) 1988-1993 The Regents of the University of California
Copyright (c) 1994 Sun Microsystems, Inc
license: ISC license
https://spdx.org/licenses/ISC.html
Sirius_Solver 6.0.0
license: EPL 2.0

Copyright (c) 2007-2018 RTE
https://spdx.org/licenses/EPL-2.0.html

Copyright © RTE 2007-2017 – Version 6.0.0
Last Rev. 25 JUN 2018

57/57



---

Source Exif Data:

File Type                       : PDF
File Type Extension             : pdf
MIME Type                       : application/pdf
PDF Version                     : 1.4
Linearized                      : No
Page Count                      : 57
Page Layout                     : SinglePage
Page Mode                       : UseOutlines
XMP Toolkit                     : XMP toolkit 2.9.1-13, framework 1.6
About                           : uuid:d9889d6d-7ab3-11e8-0000-037845dd1d30
Producer                        : PDFCreator 2.5.1.5
Keywords                        : 
Modify Date                     : 2018:06:25 11:15:49+02:00
Create Date                     : 2018:06:25 11:15:49+02:00
Creator Tool                    : PDFCreator 2.5.1.5
Document ID                     : uuid:d9889d6d-7ab3-11e8-0000-037845dd1d30
Format                          : application/pdf
Title                           : 180625-Antares-general-reference-guide
Creator                         : michdoqu
Description                     : 
Author                          : michdoqu
Subject                         : 

EXIF Metadata provided by EXIF.tools