Received April 24, 2020, accepted June 10, 2020, date of publication June 19, 2020, date of current

version July 1, 2020.

Digital Object Identifier 10.1109/ACCESS.2020.3003863

Development of a Toolbox for Alternative

Transient Program Automatic Case Creation and
Execution Directly From a Technical Database
JOÃO RODRIGO DA SILVA MUNIZ 1 , GABRIEL VIANNA SOARES ROCHA 1 ,
RAPHAEL PABLO DE SOUZA BARRADAS 2 , IGOR MEIRELES DE ARAÚJO 3 ,
DANIEL DE SOUZA AVELAR DA COSTA 1 , UBIRATAN HOLANDA BEZERRA 1 , (Member, IEEE),
MARCUS VINÍCIUS ALVES NUNES 2 , (Member, IEEE), AND JUCILENO SILVA E SILVA 4
1 Centerfor Excellence in Energy Efficiency of the Amazon (CEAMAZON), Federal University of Pará (UFPA), Belém 66077-830, Brazil
2 High and Extra-High Voltage Laboratory (LEAT), Federal University of Pará (UFPA), Belém 66075-750, Brazil
3 Laboratory of Computational Intelligence and Operational Research (LINC), Federal University of Pará (UFPA), Belém 66075-110, Brazil
4 Equatorial Pará, Electric Distribution Utility of Pará, Belém 66645-001, Brazil

Corresponding author: João Rodrigo Da Silva Muniz (rodrigosilvamuniz@gmail.com)

This work was supported in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Brazil, through the
finance code 001, in part by the Equatorial Pará, in part by the Equatorial Maranhão, and in part by the Equatorial Group.

ABSTRACT Assessing the transient behavior of electric power transmission/distribution systems, when
they are subjected to major disturbances such as lightning discharges, is an essential procedure to ensure
the operational security of these systems, in order to supply good quality power to all consumers. Electro-
magnetic transient studies are routinely performed in the planning and operation phases of these systems,
by using world widely known computer programs, among which stands out the Alternative Transient
Program (ATP), via its graphical interface ATPDraw. However, preparing files for running case studies by
using ATPDraw can become a time-consuming and human error-prone task. To overcome this drawback,
this paper presents a toolbox for automatic generation and execution of Alternative Transient Program input
files based on information directly obtained from technical databases of distribution utilities. This toolbox
aims to reduce modeling time and susceptibility to human errors in the process of describing the grid in the
software input file, which constitute the biggest obstacles to the extensive use of this program in real electric
transmission/distribution networks. For the toolbox validation, a comparison was made between the input
file generated by the ATPDraw auxiliary graphic interface and the file generated by the proposed toolbox by
using the IEEE 13-bus and IEEE 34-bus test feeders. Besides that, the simulation results of both feeders for
both softwares were compared, so that the correct operation of the Alternative Transient Program automatic
case creator and executor proposed in this article were proved. In addition, the automatic generation of ATP
cases for larger grids is exemplified with the IEEE 123-bus test feeder.

INDEX TERMS ATP, EMTP, power system transients, power system simulation, power system modeling,
power distribution.

I. INTRODUCTION arresters [2]–[4]. In low frequency phenomena, transient sim-

Studies of electromagnetic transients are essential in planning ulations can assist in assessing the impacts of inserting com-
and operation of electrical grids. The simulations of high ponents such as long extra high voltage cables, capacitor
frequency transients allow the assessment of lightning strikes banks and high voltage DC links, which can cause elec-
criticality in the electric network [1]. It indicates which tromechanical and electromagnetic phenomena with adverse
buses are more sensitive to voltage and current surges and effects on the system, such as temporary overvoltages due to
it assists in determining the optimized positioning of surge harmonic resonances [5]. Thus, the feasibility and practicality
of electromagnetic transients simulation with high detailed
The associate editor coordinating the review of this manuscript and models become of great value for energy transmission and
approving it for publication was Jenny Mahoney. distribution utilities [6].

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
114378 VOLUME 8, 2020
J. R. D. S. Muniz et al.: Development of a Toolbox for ATP Automatic Case Creation and Execution Directly

Studies on transient impacts on power systems are only the complete data description of a real distribution grid for
possible due to time-domain simulation algorithms such as transient studies using ATP is a complex and time-consuming
the Alternative Transient Program (ATP) [7]. However, elec- task, which makes it difficult to use it routinely in elec-
tric power system modeling for electromagnetic transients tric utilities’ planning and operation studies. A toolbox that
studies using specialized software, such as ATP, can be a enables the automatic generation of ATP input files directly
difficult and exhaustive task. For instance, in real electric from the electric utility’s technical database allows the mod-
distribution networks, which typically have a large number els construction easily, rapidly, and without human errors;
of electric nodes, this modeling task becomes even more it preserves the precise description of the electrical system
complicated, since a good part of these softwares requires and requires less operator intervention. Thus, studies on elec-
filling up input text files with the description of each element tromagnetic transients in large electrical networks would be
of the electric three-phase network or the insertion of these more easily executable.
elements in a graphical interface. Either way, it is necessary The simulation of electromagnetic transients in large net-
that an expert carries out the insertion of technical informa- works has been treated in the literature, whether using ATP or
tion to model manually each electric network in study. This other similar software. There is an effort to convert informa-
manual insertion takes a long time and is highly susceptible tion present in static load flow simulations or SCADA-based
to human errors, reducing work efficiency and compromising software for paid applications of electromagnetic transients
the reliability of results [5], [8]. [5], [8]. In addition, there are proprietary software created for
In order to avoid this long modeling time and reduce real-time simulation of grids [6], which prove to be efficient
the risk of human error, some possibilities are avail- but depend on high-performance simulation platforms, which
able. It is observed that most of the studies found in add financial cost to the simulation.
the literature use small stretches of grids for testing their Some articles are found in the literature regarding the
methodologies [9], [10]. This does not constitute a solution possibilities of modification or semiautomatic generation of
to the problem, but it works around it and avoids the need to ATP input files. Many works that propose ATP automation
simulate larger networks. A frequent solution is to simplify or recursive simulations depend on an ATP input file previ-
the network, which includes the use of multiport equivalents, ously modeled, either with the help of ATPDraw or directly
short circuit equivalents, high speed equivalents and line written by user, so that the tools do not automatically create
decoupling [5], [6], [8], [11]. This solution, depending on the ATP cases, only modify parameters of ATP cases already
application, can provide errors above the acceptable limit; the created [14]–[17]. Partial integration with ATP is made with
detailed modeling of the complete grid is the only possible tools such as MATLAB using batch processing mode [14],
solution [5]. [15], or specialized tools are created to carry out recursive
Specifically for the ATP software there is the ATPDraw simulation as in [18], to calculate the back flashover rate in
graphical interface, which works as a graphic preprocessor transmission lines and the insulators tolerability levels. Such
to facilitate the manual network modeling process. In this tools were also applied in calculating post-fault analysis [15]
software, the user initially creates a representative diagram and lightning discharges simulations [19].
of the system by using component models provided by ATP, In [20], developments for ATP automatic case generation
and also allows the inclusion of customized templates [12]. are presented. In this case, however, only lumped elements for
However, the insertion of each electric network’s element the transmission lines were used, and the developments relied
is a major source of complexity in the ATP utilization pro- on a database already composed by electrical parameters,
cess because it significantly increases the time for the full which is rarely found in distribution utilities. Validation is
description of the electric network and the chances of errors done by comparing it to the case created using ATPDraw,
in modeling. presenting significant errors in some cases.
Automating the manual data input procedure may have The toolbox proposed in this article aims at automating
a strong impact in reducing the time spent on the electric the generation and execution of ATP input files, facilitating
network modeling, compared to manual modeling in graph- the execution of transient studies in the routine of energy
ics software, such as ATPDraw. This reduction can allow transmission or distribution utilities, due to the fact of reduc-
the operator/engineer to focus on more productive tasks, ing the time required for grid modeling, which is a basic
such as evaluating the results obtained by simulations and step in most transient analysis of transmission/distribution
in decision-making tasks. Furthermore, errors due to human electric networks. Examples can be seen in [1], [2], [21]–[24].
disability or loss of attention can be minimized by avoiding The whole process starts by accessing a technical database
the repetition of actions [13]. whose data are similar to those usually found in the electric
All electric distribution utilities have a database with infor- distribution utilities’ databases. In a database of distribution
mation on constructive and topological description of their utilities, only geometric and topological information about
electric networks. Examples of this information are the char- the grid, such as pole height and cable spacing, are commonly
acteristics of electric cables, poles geometry, power loads, found. Direct information about electrical parameters, such
transformers, capacitor banks, surge arresters, among others. as line resistance and inductance, are missing, so it need to
Due to the large quantity of network elements to be modeled, be calculated for simulations of electromagnetic transients.

VOLUME 8, 2020 114379

 J. R. D. S. Muniz et al.: Development of a Toolbox for ATP Automatic Case Creation and Execution Directly

FIGURE 1. The proposed automatic execution tool flowchart.

With this geometric and topological basic information, the

proposed tool allows the automatic filling of the ATP input
file without needing to use the ATPDraw interface, and then,
execute the created file.
Therefore, the proposed procedure is a general tool that
FIGURE 2. Diagram of the database entities and relationships.
allows the automatic utilization of ATP directly coupled to
the technical database of electrical utilities; thus, the elec-
tromagnetic analysis of electric distribution systems using
ATP is considerably simplified. In the following sections of available in the technical databases of electric distribution
this article, the description of this toolbox will be presented utilities. Therefore, as an application dedicated to electric
in detail, as well as its validation when compared with the distribution networks, the Input Data Standardizer module of
standard ATP simulation procedure, using ATPDraw. This the proposed tool checks automatically if the electric network
validation was performed for the test feeders IEEE 13-bus topology is radial and free from islanding.
and IEEE 34-bus. After validation in two standard test grids
widely disseminated in the technical literature, the automatic A. INPUT DATA STANDARDIZATION
generation of ATP cases is exemplified in larger networks by As can be seen in Fig. 1 flowchart, the ATP cards generator
using the IEEE 123-bus test feeder. has, as input parameter, a single file, which is the output from
the Input Data Standardizer module. This file format can be
II. SYSTEM DESCRIPTION TXT, separating categories for sorting by headers, or XLSX
A normal ATP execution sequence is composed of generat- separating categories by spreadsheets. This file contains the
ing input file with extension ‘‘.atp’’ and the auxiliary files complete description of the electric network under study,
‘‘.lib’’, calling the ATP executable core, passing the input according to the following categories of network elements:
file as a parameter, and reading the output files ‘‘.lis’’ or feeder, electric node, power supply, electric branch, capacitor
‘‘pl4.’’ [25]. The input file consists of data cards, sorted into bank, transformer, load, surge arrester, electric poles config-
classes called data cases, which represent the elements and uration, electric cable type, surge arrester type, non-linear
types of elements present in the electric network, respectively, resistors, among others, and simulation settings.
according to the ATP Rulebook. Therefore, the complete The filling of this input file is based on information stored
utilization of the proposed tool needs to cover the steps of in the electric utilities technical databases, as previously men-
importing data records from a technical database, the creation tioned. An example of a typical database for this type of prob-
of a standardized input file to the cards generator, automatic lem can be seen in Fig. 2. It contains the database entity and
generation of ATP input files, ATP execution and the reading relationship diagram, where each table represents an element
of the output files. type of the electrical network, such as loads and capacitors,
The flowchart in Fig. 1 shows the sequence of steps to and each table line corresponds to a property of the element,
follow in the automatic execution of a complete simulation such as nominal power and phase sequence. Most of the
session using the proposed tool in this article. However, each information in the electric utilities technical databases con-
step operates independently and can be performed separately cerns the topological, geometric and constructive aspects of
if it is convenient for the user. the electrical network. Electrical information, as resistance,
The proposed tool was developed aiming the application in capacitance and inductance, for example, must be calculated
real electrical distribution systems, which usually have a large from these initial information stored in the technical database.
number of electrical nodes, directly using the information For instance, the registers for capacitor banks contain only

114380 VOLUME 8, 2020

J. R. D. S. Muniz et al.: Development of a Toolbox for ATP Automatic Case Creation and Execution Directly

bank equivalent capacitance, a similar procedure is adopted,

by using the nominal applied voltage and the corresponding
reactive power capacity for the capacitor bank.
p
Q = S 2 − (S · PF)2 (1)
P 2
R= 2 ·V (2)
P + Q2
Q
L = 2 · V2 (3)
P + Q2
When modeling electric transmission/distribution lines by
concentrated parameters models, the respective inductances
by phase are calculated using equations (4) and (5), adapted
from [26]. In these equations Ds is the electric conduc-
tor geometric average radius; Dij the distance between the
electric conductor and its images with respect to ground;
dij the distance between electric conductors; hi the electric
conductor average height; µ is the air magnetic permeability;
and Li the phase i inductance per unit length. Concentrated
inductance is obtained by the product of phase inductance by
the transmission line length and concentrated resistance by
FIGURE 3. Flowchart of the computer routine to check loop and island
formation in the electric network topology. the product of the conductor resistance per length multiplied
by the transmission line length.
 2h Dab Dac 
a
ln ln ln
the electric node where the capacitor bank is allocated, and 
faa fab fac
  Ds dab dac 
its nominal power. Thus, the equivalent capacitance that rep-  fab fbb fbc  = µ  D
ab 2h b D bc 

 ln ln ln


resents the capacitor bank must be calculated based on this fac fbc fcc 2π  dab Ds dbc 
basic information contained in the database.
 Dac Dbc 2hc 
ln ln ln
Before filling up the input file to the ATP cards generator, dac dbc Ds
a check is performed to verify if the electric grid is radial and (4)
1
 
has no island. These two conditions must be met, in order
to correctly operate the proposed tool. An island, for the    faa − 2 (fab + fac ) 
La
1
 
cards generator, corresponds to a set of electrical nodes that  Lb  =   fbb − (fab + fbc ) 

(5)
are not connected by any way to the main supplier source. Lc
 2 
A mesh, in turn, corresponds to the presence of more than one
 1 
fcc − (fac + fbc )
path between two electric nodes, which indicates the grid is 2
not radial. Fig. 3 presents the flowchart of the implemented The ATP card generator uses by default the JMarti
computer routine to check if these two conditions are met. distributed parameters model for representing transmis-
sion/distribution lines. This model is frequency dependent
B. CALCULATION OF ELECTRIC PARAMETERS and has a constant transformation matrix, so it is more precise
Depending on the modeling used for the electric elements for transmission/distribution lines modeling in transient elec-
representation in ATP, some parameters need to be cal- tromagnetic simulations [27]. There is a specific format for
culated previously because they are not included directly the insertion of JMarti elements in the ATP; therefore, it is not
in the database content. For distribution utilities technical necessary that the case generator calculates previously. Only
databases, these elements are capacitor banks, electric loads, in cases in which the user wants a simplified representation
and distribution lines modeled as concentrated parameters. for transmission/distribution lines, as in simulations that only
For electric loads, only the electric node where the load is involve the fundamental frequency or at electric network
connected, its apparent power in KVA, and phase sequence points not affected by the higher frequency components.
are usually available. By assuming a specified power factor, In such cases, equivalent resistance and inductance models
such as 0.92, the load active and reactive power can be are used.
obtained promptly. If the load is to be modeled by constant
impedance, for example, the equivalent parameters R (resis- C. AUTOMATIC ATP CARDS GENERATION MODULE
tance), L (inductance) and C (capacitance) to be used in The architecture of an ATP case study is formatted accord-
ATP are easily obtained by relationships between the applied ing to the pattern defined in ANSI FORTRAN 77 Stan-
voltage, and the corresponding load active and reactive power. dard [28] because that ATP software uses an installation of
The Equations used are (1), (2) e (3). For calculating capacitor EMTP FORTRAN converted from the Universal Transient

VOLUME 8, 2020 114381

 J. R. D. S. Muniz et al.: Development of a Toolbox for ATP Automatic Case Creation and Execution Directly

FIGURE 5. Example of ATP card.

FIGURE 4. ATP input file basic structure.

Program File [29]. The ATP Rule Book allows the use of input
cards with fixed format, based on the characters positioning
along the columns, or the free format, which uses tabs as
the comma [29]. The toolbox presented here uses the fixed
format, which is more explored in the ATP Rule Book and
specified by the Fortran 77 Language Reference [30]. The
input file consists of data cards ordered by categories named
as data cases. Each of these data cases matches elements
present in the electrical network. FIGURE 6. Automatic ATP cards generator flowchart.
The data cases separate different types of elements and
allocate areas in the input file for the insertion of these The automatic cards generator flowchart can be seen in
elements. The main element types are: MODELS, BRANCH, Fig. 6. The insertion order follows the data case order from
SWITCH, SOURCE and OUTPUT. The input order and the Fig. 4. It is important to note that, due to the use of the JMarti
declaration format of these elements can be seen in Fig. 4. distributed parameters models for transmission/distribution
MODELS type does not appear in all simulations, thus it lines, a prior ATP execution for each electric line is required,
is only necessary when there is use of customized elements which leads to the generation of files with extension ‘‘.lib’’
via direct programming. The BRANCH type includes linear with the electrical information of each one. In BRANCH
and non-linear elements, such as loads, transmission lines, of the ATP input file will be only the references to the files
RLC, resistance elements and nonlinear inductances. The ‘‘.lib’’.
SWITCH type includes any switch or elements disconnec-
tors. In SOURCE type are inserted the sources, either voltage D. OUTPUT FILES EXECUTION AND READING MODULE
or current source. Finally, in the OUTPUT type, the output After the ATP input file is created, a call is made to the ATP
variables are defined. Each date case is terminated with a executable core via command line by passing as parameter
BLANK statement and the full file is started and terminated the path to the created file. Thus, the simulation is performed,
with the statement BEGIN NEW DATA CASE [29]. and the output files are stored by ATP in the same folder as
Each category has its own formatting relative to the ele- the input file. The reading module finally loads the ‘‘.lis’’ file,
ment’s parameters placement on the file columns. As an which consists of a text file that has the simulation results.
example, it will be used an RLC element, which consists Each defined variable has a specific placement in the rows
of a resistor, an inductor and a capacitor in series. The RLC and columns of the ‘‘.lis’’ file. The reading module identifies
parameters are: 1  resistance; 0.1 mH inductance; and 0.01 this placement and stores the value of each iteration in vectors
µF capacitance, connected between ‘‘BUS001’’ and ground. for each variable. At the end of all lines of the file, the
This element type is positioned in the BRANCH area of the generated vectors are grouped into a key-value pair structure
file, as can be seen in Fig. 5, where the rows initiated by the whose key is the variable name.
character ‘‘C’’ indicates comments and the string ‘‘. . . ’’ indi-
cates which parts of the file have been omitted. The parameter III. RESULTS
resistance must end in column 32, and, if it is positioned The proposed toolbox validation is made through comparison
differently, it will be interpreted as a different parameter. between the results obtained by using the toolbox procedure

114382 VOLUME 8, 2020

J. R. D. S. Muniz et al.: Development of a Toolbox for ATP Automatic Case Creation and Execution Directly

FIGURE 7. IEEE 13-Bus test feeder. FIGURE 9. IEEE 123-Bus test feeder.

FIGURE 8. IEEE 34-Bus test feeder.

and the similar results obtained by running the ATP with

the ATPDraw interface. The similarity of results between
the ATPDraw simulation and the proposed toolbox results
indicates that it can be used as a choice in simulating large
FIGURE 10. IEEE 13-Bus test feeder modeled on ATPDraw.
electrical networks to minimize the time spent on modeling.
For this validation, initially the IEEE 13-bus feeder is used,
whose single line diagram can be seen in Fig. 7. This distribu-
electromagnetic analysis. These feeders modeled on ATP-
tion feeder is small and its manual modeling using ATPDraw
Draw can be seen in Fig. 10 and Fig. 11, respectively. The
is not much time-consuming. This feeder provides a severely
transient phenomenon simulated is a lightning discharge,
unbalanced system [31] and it can be useful in checking the
modeled by a Heidler current source, whose generating func-
equivalence between ATPDraw and the proposed tool. The
tion can be seen in Equation (6) and example plot can be seen
IEEE 34-bus feeder has different characteristics, being longer
in Fig. 12. The Heidler function is set as in [32].
and with more buses, as shown in the Fig. 8. The proof of
equivalence among the tools for these two different feeders  nh
ensures the functioning of ATP Cases Generator. t
I0 τ1 − t
The advantage of using the proposed toolbox in place of i (t) =  nh e τ2 (6)
ATPDraw is more evident when using larger grids, like the η 1+ t
τ1
IEEE 123-bus feeder, whose single line diagram is shown in
Fig. 9. For this feeder, the manual modeling using ATPDraw In the ATP modeling, the parameters of Table 1 were
would require a much longer time, and may require a day used. It is important to mention that voltage regulators and
or more of work, for example, for engineers to perform. branches after transformers were disregarded in both ATP-
In addition, the chance of a human error in performing the Draw modeling and the proposed tool. This modification was
task is great, since the number of elements to be manually made to obtain a feeder that has the same voltage level in all
inserted is larger. Therefore, this feeder will only be used as its extension, with the loads referenced to the high voltage
an example of the ATP Cases Generator operation, since its side. As the point to be proved is the equivalence between
equivalence to ATPDraw was proven for IEEE 13-bus and ATPDraw and ATP Automatic Case Creator and Executor,
IEEE 34-bus previously. such changes do not compromise execution.
For the IEEE 13-bus and IEEE 34-bus feeders, the results The result of steady state simulation for the IEEE 13-bus
are compared for steady state operation and for transient feeder can be seen in Fig. 13, for ATPDraw and ATP Cases

VOLUME 8, 2020 114383

 J. R. D. S. Muniz et al.: Development of a Toolbox for ATP Automatic Case Creation and Execution Directly

FIGURE 11. IEEE 34-Bus test feeder modeled on ATPDraw.

FIGURE 13. Steady state operation voltage at bus 675 of IEEE 13-Bus test
feeder.

FIGURE 12. Example of Heidler function in the form of a 10/350 lightning

current waveform with a 200kA peak [32].

TABLE 1. HEIDLER source specifications.

FIGURE 14. Transient voltage during a surge event at bus 675 of IEEE
13-Bus test feeder.

Generator. Bus 675 was chosen to exemplify the curve match- This equivalence also exists for the surge condition modeled
ing, but any other bus could be selected. It is possible to notice by a HEIDLER source, as can be seen in Fig. 14.
that both curves are completely overlapping, which indicates To ensure that ATPDraw and ATP Cases Generator present
that the two tools generate the same electrical grid. Voltages equal results, regardless of the network topology, the IEEE
at all buses in both cases were measured and plotted in the 34-bus feeder has been simulated. After running the steady
same graph and resulted in identical voltages values, which state operation condition for this feeder, both generated ATP
indicates that both tools are equivalent for the ATP creation. cases could be compared. As in IEEE 13-bus, voltages at

114384 VOLUME 8, 2020

J. R. D. S. Muniz et al.: Development of a Toolbox for ATP Automatic Case Creation and Execution Directly

FIGURE 15. Steady state operation voltage at bus 816 of IEEE 34-Bus test
feeder.

FIGURE 16. Transient voltage during a surge event at bus 816 of IEEE
34-Bus test feeder.

FIGURE 18. IEEE 34-Bus ATP case input file.

The curves overlapping with zero error indicates that the

input files describe identical systems. Fig. 18 shows parts
FIGURE 17. Zoom view of transient voltage during a surge event at of the files generated by both tools for the IEEE 34-bus
bus 816 of IEEE 34-Bus test feeder. feeder. As can be seen, except for comments, the values set
for the elements are identical. The only difference is in the
definition of the electric node’s name at which the element
all buses in both cases were plotted in the same graph and is connected. Because ATP limits the number of characters
resulted in identical values, which confirms that the two tools in the node name to only six, the proposed tool encodes grid
are equivalent for ATP case creation. Bus 816 was chosen nodes into a string as follows: the first character indicates the
to exemplify the curve matching, but any other bus could be node type, with the main buses represented by ‘‘B’’; the next
selected. The overlap of the voltage curves can be seen in four characters correspond to a hexadecimal number that is
Fig. 15. It is important to emphasize that the error between associated to a bus; the last character indicates the phase to
these curves at all buses is zero, since the generated cases are which the element is connected.
identical and the same results are expected. After validating the proposed tool by comparing it to ATP-
Similarly, after the execution of a surge condition modeled Draw for two different standard feeders, it is possible to verify
by a HEIDLER source, the voltages at all buses were mea- the advantages it provides in larger grids. As an example of
sured and plotted in the same graph. Again, the error between applying the developed tool in this type of electrical grid, the
all curve pairs was zero. The example of bus 816 can be seen IEEE 123-bus feeder was simulated in its normal operating
in Fig. 16. As the surge condition occurs during periods much condition. Voltage levels for specific buses distributed along
shorter than the sinusoidal voltage period at the fundamental the feeder can be seen in Fig. 19. The levels are within the
operating frequency, the total simulation time was reduced expected values for the nominal voltage of this grid, whose
from 20ms to 5ms. For a better visualization of the initial nominal peak value is 3.4 kV. This means that the generated
oscillations, a zoomed plot in time axis can be seen in Fig. 17. case is consistent with the expected values and, associated
In this condition, the equivalence between ATPDraw and the with the validation made in IEEE 13-bus and IEEE 34-bus
proposed tool is still preserved. feeders, the tool can be considered functional for larger grids.

VOLUME 8, 2020 114385

 J. R. D. S. Muniz et al.: Development of a Toolbox for ATP Automatic Case Creation and Execution Directly

demonstrated the great advantage as compared to the tradi-

tional ATPDraw interface; it has been estimated a time reduc-
tion from 401.3 minutes to 33.61 seconds, which corresponds
to a speedup of approximately 716 times, reducing also
chances of human errors. As illustrated by these examples,
it is evident that the proposed ATP cases automatic creator
and executor toolbox has a great advantage over the ATPDraw
traditional procedure in reducing significantly the time con-
sumed in data preparation, which is even more evident when
FIGURE 19. Voltage in phase A of electric nodes 1, 18, 54, 76, 93, 105 applying to larger electric systems.
and 150 of IEEE 123-Bus feeder.

ACKNOWLEDGMENT
TABLE 2. Case generation time comparisson. The authors acknowledge the support provided by the
Post-Graduate Electrical Engineering Program PPGEE/UFPA.

REFERENCES
[1] R. P. D. S. Barradas, G. V. S. Rocha, J. R. S. Muniz, U. H. Bezerra,
M. V. A. Nunes, and J. S. E. Silva, ‘‘Methodology for analysis of
electric distribution network criticality due to direct lightning dis-
charges,’’ Energies, vol. 13, no. 7, p. 1580, Apr. 2020. [Online]. Available:
As an estimate of time reduction provided by the proposed https://www.mdpi.com/1996-1073/13/7/1580
toolbox, the time spent in modeling the IEEE 13-bus and the [2] G. V. S. Rocha, R. P. D. S. Barradas, J. R. S. Muniz, U. H. Bezerra,
IEEE 34-bus feeders on ATPDraw was 42 and 112 minutes I. M. D. Araújo, D. D. S. A. D. Costa, A. C. D. Silva, M. V. A. Nunes,
and J. S. E. Silva, ‘‘Optimized surge arrester allocation based on
of uninterrupted work, respectively. The ATP Cases Gener- genetic algorithm and ATP simulation in electric distribution systems,’’
ator modeled the same feeders in 7.14 and 16.8 seconds, Energies, vol. 12, no. 21, p. 4110, Oct. 2019. [Online]. Available:
respectively, being at least 353 times faster than the ATPDraw https://www.mdpi.com/1996-1073/12/21/4110
modeling. It is possible to see a comparison of the total time [3] E. Bullich-Massague, A. Sumper, R. Villafafila-Robles, and J. Rull-Duran,
‘‘Optimization of surge arrester locations in overhead distribution net-
and time per bus for the case generation of the three feeders works,’’ IEEE Trans. Power Del., vol. 30, no. 2, pp. 674–683, Apr. 2015.
simulated in the Table 2. For the IEEE 13-bus and IEEE [Online]. Available: http://ieeexplore.ieee.org/document/6783697/
34-bus feeders, which were modeled in both software, the [4] R. Shariatinasab, H. Falaghi, and J. Ghayur Safar, ‘‘Optimisation of
arrester location in risk assessment in distribution network,’’ IET Gener.,
average time per bus when using ATPDraw is 195.75 seconds. Transmiss. Distrib., vol. 8, no. 1, pp. 151–159, Jan. 2014. [Online].
If this average time is maintained for IEEE 123-bus, the total Available: https://digital-library.theiet.org/content/journals/10.1049/iet-
estimated time will be 24077.25 seconds, or approximately gtd.2013.0166
[5] S. Dennetière, A. Parisot, E. Milin, and A. D. Pons, ‘‘Resonance
6.7 hours of uninterrupted work, disregarding human delays and insertion studies with EMTP: Working with large scale network
and errors. The ATP Cases Generator performs the same task models,’’ in Proc. Int. Conf. Power Syst. Transients (IPST), 2011,
in 33.61 seconds, approximately 716 times faster. The com- pp. 1–6. [Online]. Available: https://ipstconf.org/papers/Proc_IPST2011/
11IPST087.pdf
putational environment used to run the ATP Cases Generator [6] W. Weiwei, Z. Yiying, L. Chong, D. Peng, H. Tao, L. Baiqing,
and ATPDraw has an Intel Core i5 9400F processor with a L. Yalou, H. Fei, and Z. Yi, ‘‘An implementation technology of elec-
frequency of 2.90 GHz and 24 GB of DDR4 RAM (3×8 GB) tromagnetic transient real-time simulation for large-scale grid based
on HYPERSIM,’’ in Proc. Int. Conf. Power Syst. Technol. (POWER-
with a frequency of 3600 MHz. The version of ATPDraw used CON), Nov. 2018, pp. 167–172. [Online]. Available: https://ieeexplore.
was 5.9p4 for Windows. The proposed toolbox presented ieee.org/document/8602121/
consistent results for a larger grid, which ensures significant [7] H. W. Dommel, ‘‘Digital computer solution of electromagnetic transients
in single-and multi-phase networks,’’ IEEE Trans. Power App. Syst.,
time savings in simulations that will be used at the operational vol. PAS-88, no. 4, pp. 388–399, Apr. 1969.
level in distribution utilities. [8] P. Zhao, J. Yang, M. Guo, F. Fei, R. Huang, A. Pang, M. Yang,
and Q. Huang, ‘‘Rapid modeling of large-scale power system based
on PSCAD,’’ in Proc. IEEE Asia Power Energy Eng. Conf.
IV. CONCLUSION (APEEC), Mar. 2019, pp. 167–171. [Online]. Available: https://
In this work, a toolbox for automatic creation and execution ieeexplore.ieee.org/document/8720692/
of ATP cases was proposed. The toolbox uses the distribu- [9] A. L. Orille-Fernandez, N. Khalil, and S. Rodriguez, ‘‘Failure
risk prediction using artificial neural networks for lightning surge
tion utility’s technical database directly and performs calcu- protection of underground MV cables,’’ IEEE Trans. Power Del.,
lations and transcriptions necessary for formatting accord- vol. 21, no. 3, pp. 1278–1282, Jul. 2006. [Online]. Available:
ing to the ATP accepted standard. The comparison of the http://ieeexplore.ieee.org/document/1645166/
results obtained by the ATPDraw auxiliary graphic software [10] M. Omidiora and M. Lehtonen, ‘‘Mitigation of lightning flashover from
tree to medium voltage aerial cable using shield wire,’’ IEEE Trans.
and the proposed ATP cases automatic creator and executor Power Del., vol. 32, no. 4, pp. 1924–1934, Aug. 2017. [Online]. Available:
for the IEEE 13-bus and IEEE 34-bus systems indicated http://ieeexplore.ieee.org/document/7500065/
the complete equivalence between them, with zero error in [11] J. Xu, Y. Zhao, C. Zhao, and H. Ding, ‘‘Unified high-speed EMT equivalent
and implementation method of MMCs with single-port submodules,’’
all variables. After the validation of the proposed tool, the IEEE Trans. Power Del., vol. 34, no. 1, pp. 42–52, Feb. 2019. [Online].
creation of the IEEE 123-bus case study using this toolbox Available: https://ieeexplore.ieee.org/document/8489923/

114386 VOLUME 8, 2020

J. R. D. S. Muniz et al.: Development of a Toolbox for ATP Automatic Case Creation and Execution Directly

[12] A. C. Marchesan, L. Mariotto, F. Loose, P. F. Alves Fo, G. Cardoso, [32] B. R. Terespolsky, ‘‘An approximation to the heidler function
and M. C. Camargo, ‘‘ATP substation analysis: Graphical preprocessor with an analytical integral for engineering applications using
software for transient analysis in substations,’’ in Proc. 50th Int. Univ. lightning currents,’’ M.S. thesis, Dept. Eng. Built Environ., Univ.
Power Eng. Conf. (UPEC), Sep. 2015, pp. 1–6. Witwatersrand, Johannesburg, South Africa, 2015. [Online].
[13] L. Argote, Organizational Learning: Creating, Retaining and Transferring Available: http://wiredspace.wits.ac.za/jspui/bitstream/10539/22605/2/
Knowledge. New York, NY, USA: Springer, 1999. [Online]. Available: BrettTerespolskyMScDissertation.pdf
https://books.google.com.br/books?id=4m_I1oBnfOoC
[14] A. Abdullah, ‘‘ATPMAT: An open source toolbox for systematic cre-
ation of EMTP cases in ATP using MATLAB,’’ in Proc. North
Amer. Power Symp. (NAPS), Oct. 2015, pp. 1–6. [Online]. Available: JOÃO RODRIGO DA SILVA MUNIZ was born in
https://ieeexplore.ieee.org/document/7335090 Belém, Brazil, in 1997. He received the degree in
[15] R. Niu, J. Liang, Z. Yun, H. Wang, and Z. Wei, ‘‘A simulation toolbox electrical engineering from the Federal University
based on an integrated automatic mechanism for ATP and MATLAB,’’ in of Pará, in 2019, where he is currently pursuing
Proc. IEEE Int. Conf. Mechatronics Autom., Aug. 2014, pp. 2039–2043. the master’s degree in electrical engineering in the
[Online]. Available: http://ieeexplore.ieee.org/document/6886017/
area of energy systems and the field of research in
[16] S. Su, X. Duan, and X. Zeng, ‘‘ATP-based automated fault simulation,’’
electric power systems.
IEEE Trans. Power Del., vol. 23, no. 3, pp. 1687–1689, Jul. 2008. [Online].
Available: http://ieeexplore.ieee.org/document/4529108/ He is a Researcher of the Research and Devel-
[17] J. J. Mora, J. C. Bedoya, and J. Melendez, ‘‘Extensive events opment Project Optimized Surge Arrester Alloca-
database development using ATP and MATLAB to fault location in tion in Distribution Grids, which is carried out in
power distribution systems,’’ in Proc. IEEE/PES Transmiss. Distrib. partnership with the utilities Equatorial Pará and Equatorial Maranhão.
Conf. Expo., Latin Amer., Aug. 2006, pp. 1–6. [Online]. Available:
https://ieeexplore.ieee.org/document/4104657
[18] L. Hailiang, W. Xishan, L. Lei, and T. Xuefang, ‘‘Automatic simulation
of transmission line backflashover rate base on ATP,’’ in Proc. GABRIEL VIANNA SOARES ROCHA was born
3rd Int. Conf. Electr. Utility Deregulation Restructuring Power in Belém, Brazil, in May 5, 1992. He received
Technol., Apr. 2008, pp. 1211–1215. [Online]. Available: https:// the B.Sc. degree in computer engineering from the
ieeexplore.ieee.org/document/4523591 Institute of Superior Studies of Amazon (IESAM),
[19] W. Castro and J. A. Vasconcelos, ‘‘Optimized positioning of surge arrester Brazil, in 2014, and the M.Sc. degree in elec-
in transmission lines using genetic algorithm and ATP,’’ in Proc. 21st trical engineering with an emphasis in electric
Brazilian Congr. Autom. Vitória, Brazil: Brazilian Society of Automation, power systems from the Federal University of
2016, pp. 1–6. Pará (UFPA), Brazil, in 2017, where he is cur-
[20] J. R. S. Muniz, J. A. A. Moraes, G. V. S. Rocha, M. V. A. Nunes, rently pursuing the Ph.D. degree, through the
R. P. S. Barradas, U. H. Bezerra, A. B. Brito, F. P. Monteiro, and Post-Graduate Electrical Engineering Graduate
R. L. S. Carvalho, ‘‘ATP cards automatic generation from an electrical Program (PPGEE), with an emphasis in electric power systems.
network elements database using Python,’’ in Proc. 13th IEEE Int. Conf.
In 2013, he was affiliated with the Center for Excellence in Energy
Ind. Appl. (INDUSCON), Nov. 2018, pp. 564–570. [Online]. Available:
Efficiency of the Amazon (CEAMAZON), UFPA. His research interests
https://ieeexplore.ieee.org/document/8627278/
[21] M. Davoudi, J. Sadeh, and E. Kamyab, ‘‘Transient-based fault loca- include computational intelligence and optimization algorithms applied to
tion on three-terminal and tapped transmission lines not requiring line electric power systems.
parameters,’’ IEEE Trans. Power Del., vol. 33, no. 1, pp. 179–188,
Feb. 2018.
[22] S. Robson, A. Haddad, and H. Griffiths, ‘‘A new methodology for
network scale simulation of emerging power line communication stan- RAPHAEL PABLO DE SOUZA BARRADAS was
dards,’’ IEEE Trans. Power Del., vol. 33, no. 3, pp. 1025–1034, born in Pará, Brazil, in 1990. He received the
Jun. 2018. bachelor’s and master’s degrees in electrical engi-
[23] M. Arkan, D. Akmaz, M. S. Mamiş, and M. E. Tağluk, ‘‘Design and neering from the Federal University of Pará, Pará,
development of travelling-wave-frequency-based transmission line fault in 2015 and 2017, respectively, where he is cur-
locator using TMS320 DSP,’’ IET Sci., Meas. Technol., vol. 13, no. 4, rently pursuing the Ph.D. degree.
pp. 518–522, Jun. 2019. He develops studies related to electromagnetic
[24] M. A. Shokry, A. Khamlichi, F. Garnacho, J. M. Malo, and F. Alvarez, transients in distribution networks, arresters, and
‘‘Detection and localization of defects in cable sheath of cross-bonding high-voltage tests and grounding.
configuration by sheath currents,’’ IEEE Trans. Power Del., vol. 34, no. 4,
pp. 1401–1411, Aug. 2019.
[25] L. Prikler and H. K. Hoidalen. (2009). ATPDraw Version 5.6 for
Windows 9x/NT/2000/XP/Vista Users’ Manual, [Online]. Available:
http://www.elkraft.ntnu.no/atpdraw/ATPDMan56.pdf IGOR MEIRELES DE ARAÚJO was born in
[26] M. V. Bakshi, Electrical Power Transmission And Distribution. Belém, Brazil, in March 1, 1992. He received the
Pune, India: Technical Publications, 2007. [Online]. Available: B.Sc. degree in computer engineering from the
https://books.google.com.br/books?id=REww2ZF2RwwC Institute of Superior Studies of Amazon (IESAM),
[27] J. Marti, ‘‘Accurate modelling of frequency-dependent transmission lines Brazil, in 2014, and the M.Sc. degree in applied
in electromagnetic transient simulations,’’ IEEE Trans. Power App. Syst.,
computing for electrical engineering from the Fed-
vol. PAS-101, no. 1, pp. 147–157, Jan. 1982.
eral University of Pará (UFPA), Brazil, in 2017,
[28] The Fortran Company. (1978). FORTRAN 77 Standards. [Online].
Available: https://wg5-fortran.org/ARCHIVE/Fortran77.html where he is currently pursuing the Ph.D. degree,
[29] A. Rifaldi and B. Lastra, Electronic Edition of the ATP Rulebook in PDF through the Post-Graduate Electrical Engineering
Format. La Plata, Argentina: Revista Iberoamericana del ATP, 2001. Graduate Program (PPGEE), with an emphasis in
[30] Sun Microsystems. (2001). FORTRAN 77 Language Reference. applied computing.
[Online]. Available: https://docs.oracle.com/cd/E19059-01/fortec6u2/ In 2014, he was affiliated with the Laboratory of Computational Intel-
806-7986/806-7986.pdf ligence and Operational Research (LINC), UFPA. His research interests
[31] IEEE Recommended Practice for Electric Power Distribution System Anal- include computational intelligence, data mining, and parallel computing
ysis, Standard 1729-2014, 2014, pp. 1–20. using GPGPU.

VOLUME 8, 2020 114387

 J. R. D. S. Muniz et al.: Development of a Toolbox for ATP Automatic Case Creation and Execution Directly

DANIEL DE SOUZA AVELAR DA COSTA was MARCUS VINÍCIUS ALVES NUNES (Member,
born in Belém, Brazil, in 1995. He is currently IEEE) received the B.E. and M.E. degrees in elec-
pursuing the degree in computer engineering with trical engineering from the Federal University of
the Federal University of Pará. He has experi- Pará (UFPA), in 1994 and 1996, respectively, and
ence in computer science, focusing on analytical the Ph.D. degree in electrical engineering from
and simulation models. He is a Researcher of the Federal University of Santa Catarina, in 2003.
the Research and Development Project Optimized He was the Coordinator of the Post-Graduate Elec-
Surge Arrester Allocation in Distribution Grids, trical Engineering Graduate Program (PPGEE),
which is carried out in partnership with the utilities ITEC, UFPA, from November 2008 to November
Equatorial Pará and Equatorial Maranhão. 2012. He is currently an Associate Professor IV
with the UFPA, where he is also the Coordinator of the High and Extra-
High Voltage Laboratory and CNPq High Voltage Research Group. He is the
Vice-Chair of the Institute of Electrical and Electronics Engineers (IEEE)
North Central Section of Brazil.
UBIRATAN HOLANDA BEZERRA (Member,
IEEE) was born in Pereiro, Brazil, in 1950.
He received the degree in electrical engineer-
ing from the Federal University of Pará (UFPA),
in 1976, the master’s degree in electrical engi- JUCILENO SILVA E SILVA was born in Castanhal,
neering from the Federal University of Itajuba, Brazil, in 1977. He received the Eng. and M.Sc.
in 1980, and the Ph.D. degree in electrical engi- degrees in electrical engineering from the Federal
neering from the Federal University of Rio de University of Pará (UFPA), Belém, Brazil, in 2002
Janeiro, in 1985. and 2006, respectively, where he is currently pur-
He is currently a Professor with the UFPA, suing the Ph.D. degree in electrical engineering.
working in undergraduate and postgraduate education, with the master’s Since 2007, he has been an Electrical Engi-
degree and doctorate guidance. He is a Researcher with the Center for neer with Environmental and Expansion Planning
Excellence in Energy Efficiency of the Amazon (CEAMAZON), UFPA, Corporate Management, Equatorial Pará, Electric
where he develops research and consultancy for electrical utilities, focusing Distribution Utility of Pará, where he develops
on preferred themes such as power systems security assessment, power electrical system analysis studies for the expansion of sub-transmission and
quality, distributed generation, the applications of computational intelligence distribution networks. His research interests include electric power distribu-
techniques in power systems, and the planning and operation of electric tion and electromagnetic transients.
distribution grids.

114388 VOLUME 8, 2020