Developing Applications for State Estimator DIgSILENT Marcos Sacasqui

Paper Title (use style: paper title)

Developing applications for State Estimator DIgSILENTUsing Microsoft Excel As DatabaseMarcos R. SacasquiPostgraduate Student University of [email protected] In order to estimate the state of electric power systems: power measurements. In cases of study or revision is necessary to use a DPL program (DIgSILENT Language Program) to perform these tasks continuously. This helps in identifying errors in the measurements and the results shared in programs like Microsoft Excel.

Keywords State Estimation, Observability Check, Plausibility Check, Measurements, Simulation, Electric Power Systems, Macros, Procedures, Forms.I. INTRODUCTIONIn all control center world energy state estimator tool (ES) is needed [4]. In Peru DIgSILENT used to study the simulation of power systems, however, it is necessary to develop a complementary program that allows you to update your database (External Measurements); This requires knowledge of the algorithm and programming. The use of DPL facilitate the automation of various tasks, being Object-oriented programming, it is possible to use Standard models simulation grids (Electric Networks). Once the programming is complete DIgSILENT to exploit results are used to DDE (Dynamic Data Exchange) to establish a connection with Microsoft Excel where you can easily make presentations, graphics, comparison and creating reports using Visual Basic: Forms, Modules: Functions and procedures.II. OVERVIEW STATE ESTIMATOR DIgSILENT

The estimation tool to DIgSILENT is an application that uses an iterative Newton-Raphson, where the approach to solving the minimization problem is based on Lagrange multipliers. If the observability analysis is correct, the convergence (in general) is guaranteed. [3]Today many control centers need a tool Estimation status that allows comparisons to verify information obtained from the principal sources of information, an application that can develop in DPL is the automation of this tool to generate reports in Excel format and also update your database from Excel file, remember that DDE is enabled and excel has great benefits when using multiple data connections (ODBC). It is for this reason that we made the application of state estimation through iterative algorithms implemented in DIgSILENT.For any case, the advantage of working with the software DIgSILENT in Peru is that it can easily increase the reliability of the analysis based on the information added by the control center operators, also using the advantage of standardization of this Software in Peru which facilitates the comparison of results between companies and Coordinators.A. Defining Procedures and Objective Flowchart:

Fig. 1. Representation of State Estimation en DIgSILENT [1]Objective function (1): Mathematically, this can be expressed with a weighted square sum of all deviations between calculated values (calVal) and measured (meaVal)

x: Vector containing the variables to calculate. wi: Matrix based on the standard deviation of the measurement instruments Weighting.

Testing results (plausibility test): It is intended to detect and remove all measurements with some apparent error. Power Factory provides the tool that allows us to control our measurements using excess factors to address active power flow, branches Lost, Lost in passive elements, flow in open branches, nominal values on the branches, Sum nodes P, Q.

Observability: Generally speaking, a region of the network is called observable, if the system measurements provide sufficient information to estimate the state of that part of the network. The method used is "raking sensitivity matrix"

Detection of bad data: The error of each measuring device can be estimated by evaluating the difference between the calculated and measured amount. Extremely distorted measurements (e.g., the estimated error is much greater than the standard deviation of the measuring device) are not considered in the subsequent iterations. The process is repeated until there are no bad measurements. We can assign up to eliminate bad measurements.

Loop state estimation: To improve results, observability data and bad-detection Objective Function (minimization) must iterate.

III. SETTINGS ES DIgSILENT Variables to be added to the grid (notation using DIgSILENT): (External) P-Measurement (StaExtpmea) (Active Power Measurement) [1] (External) Q-Measurement (StaExtqmea) (Measurement of Reactive Power) [1] (External) I-Measurement, current magnitude (StaExtimea) (measurement variable current) [1] (External) V-Measurement, voltage magnitude (StaExtvmea) (Measuring Voltage Magnitude) [1] (External) signalization Breaker Breaker Status (StaExtbrkmea) (State On Off Switch) [1] (External) Tap-Tap Position Measurement Position (StaExttapmea) (Position Tap State) [1]Any number of measurements may be defined in the cubicle. The location of the gauges is in strategic cubicles:

Fig. 2. Adding external measurements

Fig. 3. Selecting the external measurement

Fig. 4. Completing information required by Active Power Meter

IV. IMPLEMENTATION OF STATE ESTIMATOR FOR ELECTRIC POWER SYSTEM EGASAThe central challenge in the operation of electric power systems is often cited: Electricity must be generated in the exact moment that it is consumed [6]. To move electric power through a network is to obey the law of energy conservation; what goes in must come out. Unlike a natural gas pipeline that can accommodate a variation in gas pressure and thus serve simultaneously as conduit and storage reservoir, a transmission line cannot store electricity. [2][5] Changes in this equation make power measurements on nodes vary. A system to review the set of measurements is necessary, so State Estimator is used.Measurements are obtained from SCADA systems, which also provide:Network topology: The graphs of networks to allow entry DIgSILENT Impedance for passive and active network elements. Enables the rapid creation of substation bays where you can modify the states of an electrical network switches.

Operation Data: Corresponding to the time at which the simulation is carried out, remember that talk of a simulation that is not dynamic. Among these we consider the Demand and Supply of Electrical Diagram load profile generation.

Meters: As the measurements that we can deliver the SCADA be necessary to update the Network using state estimator, to execute the algorithm without errors is advisable to put gauges on the nodes listed, however this is not strict.

V. Developing DPL application for state estimator

The DPL application is a type of object-oriented programming called, the same that uses Dynamic Data Exchange (DDE) to interact with MICROSOFT EXCEL. Excel has Visual Basic (VBA) to automate tasks or graphical comparisons

In this document show appropriate programming to update values in different external measurements and developments relating to Excel. Keep in mind that the grids may vary according to the implementation of national transmission plan. The program must be filtered by type and through all the existing elements to record information correctly.

Fig. 8. DDE Connection to MS EXCEL

Fig. 9. DIgSILENT database update data in this case for reactive power measurements.

Fig. 10. Running ES algorithm from DPL

Fig. 11. Implementation forms Visual Basic (Obtaining Data from DIgSILENT)

Fig. 12. Creating database in Excel Macros review (Obtaining Data from DIgSILENT)

Fig. 13. Form of alarms (VBA Excel)

Fig. 14. Active Power measurements at the points provided. Analysis in EGASA-2015 network.VI. CONCLUSIONS

It is possible to interact DIgSILENT-EXCEL (Exchange information in a control center and analyzing) to create custom programs that allow us to check data of measurements. In Figure 14 we can observe the proper performance of state estimator in the electric power system EGASA. The state estimation algorithm in network performance EGASA, according to the Table 1 we have 0.244% maximum error between estimated and actual measurements.Table. 1. Real measurements and Estimated measures in transmission lines. (L3091 with disturbance)Real MeasurementsES( With Disturbances)

Point Of MeasureP. Act. [MW]P React. [MVAR]P. Act. [MW]P React. [MVAR]

L 3061 5/94.7218240.82391754.7234870.8236759

L 3060 8/911.886893.25800511.888553.257776

L3072 9/10-4.766125-1.125842-4.766929-1.125849

L 3071 9/1-8.592718-2.365523-8.594131-2.365482

L 3080 9/1-8.592718-2.365523-8.594131-2.365482

L 3100 2/35.241442.2095335.2396462.209253

L 3101 12/25.439021.3871735.4369681.387103

L 3091 1/2-5.714188-0.170719-5.719242-0.1705771

L 3090 1/2-5.714188-0.170719-5.719242-0.1705771

L 3101(1) 12/3-4.958999-2.903275-4.957507-2.902876

L 3061(2) 3/40.66926680.99006930.66763450.9903543

L 3060(2) 3/62.4547621.5794832.4531281.579757

L3072(1) 11/112.130873.80450512.132633.804562

About of the State Estimation Simulation: sensitivity matrix 41x32 dimensin, pseudorank is 32. Newton-Raphson Iteration converged successfully in 2 Iterations. Total of 13 out of 79 measurement(s) are declared to be bad P-measurement: 7(39); Q- measurement: 5(39); I- measurement: 1(1), In Visual Basic detect (5) abnormal measurement values, which are shown in Figure 13.VII. AcknowledgmentWe thank EGASA, Per, for providing the network specifications of their model and data of Operation.VIII. References

[1] PowerFactory Users Manual DIgSILENT PowerFactory Version 14.0. DIgSILENT GmbH Gomaringen, Germany, pp. 34-1, July 2010. [2] Von Meier Alexandra, Electric Power Systems - A Conceptual

Introduction, PowerSystem Analysis. IEEE PRESS, Wiley, 2006[3] Jochen Alber, State Estimation in PowerFactory: Algorithmic Aspects RTE-VT workshop, Paris, May 29-30, 2006[4] Abur Ali, Gmez Expsito Antonio, Power System State Estimation: Theory and Implementation CRC Press 2004[5] Kothari D.P. y Nagrath I.J., Sistemas elctricos de Potencia McGraw-Hill Education, 2008[6] Saadat Haadi., Power System Analysis McGraw-Hill Education, 2002(1)

Fig. 5. Electric Power System EGASA: Interconnection is modeled as Slack bus (32), 138kV lines in bars Callali (1) and Yura (5). The numbers indicate that nodes Power meters are located.

Fig. 6. Interface between the control scheme and the State Estimator DIgSILENT.

Fig. 7. The benchmark network.

_1491830671.vsd

SERVICIO EXCEL ACTUALIZADO POR ODBC

Tpico WorkBook

BASE DE DATOS ORACLE

Fuente: SCADA

Consulta Automatica

CONEXIN DDE PF DIgSILENT(DPL)

Tpico WorkBook

Continua Actualizacion de Datos

Comparacin ES/PF VS Mediciones