Topology processing detects the supply status of individual network circuits based on the current states of switches and potential points. The topology is defined during graphical data model entry by deploying technological operating equipment during a work step. The possible supply states are distinguished through colour on the process screens as being e.g. de-energized or earthed. Topology colour inheritance from transformers or bus bars may optionally be differentiated according to network groups or network lines and circuits.
After an earth fault has occurred, the system evaluates the current states of the earth fault direction relays, narrows down the point of the fault as far as possible and visualises the results. The indicated directions of the relays can, for example, be shown as arrows along the lines. The direction relays are reset automatically.
Switching programs make it possible to configure and test standard switching operations (e.g. feeder ON/OFF), including the necessary interlocking conditions, whilst taking operating times into consideration; once configured and tested, these switching programs can then be activated whenever needed for network operations. Frequent switching programs, e.g. for switching a standardised bay, are defined once and can then be applied to all bays of the same type and with the same interlocking conditions.
Switching programs can be initiated on an event driven, scheduled or manual basis. They can be halted, continued or aborted at any particular point. Each switching step is checked for success; if it fails, the program is aborted.
A forecast is used to predict the development of a dimension which cannot be determined, i.e. to forecast the load development over a certain period of time, taking into account factors such as the type of day, meteorological data, etc. The use of artificial neural networks (ANN) has proven to be successful for these types of applications, as they are capable of supplying very precise results very quickly. Solutions based on artificial neural networks assume that the progressions known from the past represent regularities; even though these regularities cannot be explicitly defined, it is generally accepted that they are valid and can be projected into the future. The future progression is derived with excellent precision from regularities implicitly coded into old data and from the current and estimated future auxiliary energising values, whereby the forecasts adapt themselves to gradual changes in consumption characteristics and load behaviour.
This means that a forecast method based on artificial neural networks is a system which, through training, is able to forecast the time-related progression of a variable using known data. This method has proven to be superior compared to conventional algorithm-based methods.
Incorporated into the ControlStar system, the PROGNOS software module is universally applicable and is specially adapted to forecasting load developments (electricity, gas, district heating, etc.).
The aim of optimisation is to achieve uniform power supply while complying with the supply maximum in each billing period through:
load transfer
consumption limitations for bulk consumers
use of peak-load installations
use of energy storage facilities or other generators.
Optimisation is achieved using interchange power control. There are three different billing periods, i.e. 15, 30 and 60 minutes. The billing period signal – usually provided by the upstream supplier – is monitored and reproduced in the event of a failure. The optimisation calculation function determines the power available up to the end of the billing period and compares this with the current power consumption. If there is a difference, load switching is employed in an attempt is made to correct the discrepancy.
Before switching a load, various tests are carried out, e.g.:
Switching status, priority, nominal power
Min. and max. ON and OFF times as well as switching frequency
Run-up and run-on time
The load characteristics are entered in the data model, which can be modified online. Optimisation can be visualised using:
Status overview for loads
Displays of the current, correction and free power (in both numeric and graphic form)
Ripple control systems are used for load management in electrical networks. These systems are used to execute long-term, cyclically recurring or spontaneous switching actions derived from load development in the network. Connecting or disconnecting generators or loads, depending on the time or the current load situation, guarantees contractual compliance and influences load development, meaning the network can be operated more economically.
Load control scheduling is a ControlStar system function that can either replace a conventional, separate ripple control command unit or be used in parallel.
Ripple control commands can be triggered by:
Time lists for planned switching actions, also taking the type of day into consideration
Manual control
External events
Load dependent management (cost optimization for electricity)
Visualization is carried out using:
Status overviews of ripple-control commands, transformer substations, external events and time-meter states
The actual state of the network may be determined using network safety analyses, which analyses possible faults as a preventative measure. Load flows and losses in the network resulting from planned switching states and load situations are also calculated.
The load profile calculation function is employed to calculate a simulated network state. This function requires breaker position signals and transformer taps as well as feeding and load values as input variables, and uses these values to calculate node voltages, flows, currents and losses. Warnings are generated in the event of overloads, voltage band violations or when the permitted limits for the reactive power of voltage regulating generators have been reached. The input variables can be taken from the result data record of the state estimation function or from an archived process data screen. The load profile program is based on the Stott decoupled load profile approach. A Newton-Raphson load profile program is used in special cases, such as city grids and industrial networks with high cable content.
The switching simulation function can be reached from the equipment switching dialog fields, making it quick and convenient to use. The desired switching status is simulated and the load profile – together with all node voltages, currents and losses – is calculated using the load profile calculation function. After verifying the simulation result, the switching status can be released for execution.
The contingency analysis function calculates and monitors the (n-1) contingency of the network. A list of failure variants (e.g. line or bus bar failure) is generated automatically; however a manually updated list and a list of field equipment subject to an absolute and relatively high load are also taken into consideration. Warnings are issued for failure variants that lead to overloading or network failures. It is also possible to assume that several items of field equipment have failed.
The short circuit calculation function allows the permissible short circuit power and currents in the network to be monitored. Two methods are provided to calculate symmetric (three phase) and asymmetrical short circuits according to VDE 0102.
In the first method calculation can be carried out for all short circuit points (bus bars) in a single pass. The short circuit variables are calculated for the individual fault points as well as for nearby equipment, and are compared with the permissible maximum values provided by the corresponding circuit breakers.
With the second method, a short circuit point is specified at any particular bus bar in the network. The resultant short circuit currents and power levels in all the network branches, the short circuit voltages at all bus bars as well as the decay coefficient, steady short circuit current, etc. are calculated. In the case of asymmetrical short circuits, the currents and voltages are output for the individual phases.
The equivalent network calculation function performs the online reduction of external networks that cannot be monitored, i.e. are considered in the form of equivalent branches and feedings. If, for instance, part of the company’s own network cannot be monitored due to measurement failures, this area can be replaced using the online equivalent network calculation function.
In small companies, in particular, the network control room is usually only manned during the day. It may be necessary to call up standby personnel in the evening and at night in order to immediately locate network faults and to eliminate these as soon as possible. In this case, the control system can automatically activate terminals, such as telephone, fax, city call or company radio transmission.
Communication with the control system is managed via SMS, mail or synthetically generated speech messages. ControlStar is equipped with suitable functions for planning and administration of standby duties.
Topology recognition