Loss Minimization (LM)

Design Features

The Loss Minimization (LM) function is designed to automatically monitor and control individual switchable capacitor banks in order to minimize overhead feeder losses. Feeder loss minimization is realized by reducing reactive power flows while maintaining voltages and power factors within specified limits. The LM is a real-time function and executes periodically at a dispatcher adjustable time interval.


Using capacitor bank control LM function requires a control interface to the SCADA system via RTUs. By means of SCADA facilities, LM analyzes prevailing substation and local conditions and, as necessary, switches the individual capacitor banks of each feeder on or off in order to minimize the feeder's losses. This is performed while maintaining the LM calculated capacitor bank voltages and power factors within specified limits. LM coordinates its operations with the status of substation load-tap-changing transformers (LTCs) and capacitor banks, although these devices may not be controlled by the SCADA system.


Loss Minimization executes periodically at a dispatcher adjustable time interval (typically every 30 minutes) and alarms the dispatcher whenever a switch fails to operate. LM does not allow unbalanced switching of a capacitor bank to take place if an individual switch should fail; a failed switch causes the capacitor bank to be unavailable to LM. In addition, LM maintains a count: of capacitor bank switch operations and will alarm the dispatcher when the number of counts exceeds a specified maximum over a specified time interval.


The dispatcher has the ability to enable and disable LM monitoring and control of individual capacitor banks as well as the ability to activate and de activate LM. For individual capacitor banks, the dispatcher also has the ability to assign voltage limits, adjust switch operation count rates (maximum counts and time intervals), and reset accumulated switch operation counts to zero.


In the LM function, individual capacitor banks are processed on per feeder basis. Similarly, the ON/OFF control commands to capacitor banks are also grouped based on each feeder. Capacitor banks that belong to a particular feeder are identified through network tracing by using network connectivity and dynamic switch status information.


In the process of capacitor bank control, individually operable capacitors on the feeder are identified by topology tracing from a feeder breaker downstream. Feeder loads are estimated to calculate voltage, branch flows, and power factors. The branch flows at capacitor locations are analyzed to find those capacitors whose branch reactive power exceeds a pre-determined limit. These capacitor banks are sorted in descending order based on their branch reactive power flows. The capacitor with the largest branch reactive power is selected as a LM control candidate. Its impact on voltage and power factor at the capacitor location is calculated and checked against their corresponding limits. If any one of the two constraints is violated, this capacitor bank will be passed over and the next capacitor is processed. Otherwise, a control command is issued to operate this capacitor bank. To prevent unbalance switching of a capacitor bank, a verification procedure will follow to check if any switch operation has failed. If so, this capacitor bank will be disabled for LM function. The above process is repeated until no capacitor is found whose branch reactive power exceeds the limit.


The LM function is designed to operate under the following assumptions: 

  • Distribution system is in a radial configuration; no loops are involved within the LM control area.
  • Each load and capacitor bank can be represented as constant P and Q and/or constant current injections.
  • Existing real-time condition is used as the base point for load flow analysis. 


The LM function is also subject to the following constraints:

  • High and low voltage limits at capacitor bank locations
  • Power factor limits at capacitor locations
  • Maximum number of operations per capacitor bank during specified time period
  • Operability of capacitor banks


The main task of the LM function is to identify capacitor bank locations, to determine whether it is necessary to operate capacitors, and to make sure such operations do not cause any constraint violation. The first part of the LM function is accomplished through topology tracing process. To determine when and how to operate a capacitor, the following rule is applied: A capacitor needs to be operated if there exists a significant amount of lagging reactive power in the upstream branch flow. Likewise, it needs to be turned off if there is a large amount of leading reactive power in the branch flow. The decision can be made through a series of load flow calculations. Finally, to verify if a given capacitor operation violates any voltage or power factor constraints, the changes in voltage and power factor are calculated considering the effect of the capacitor operation.

In the test feeder depicted in the graphs below, the KVAr loss minimization represented a savings of $43,500 per year based on a price of $0.06/KWHr. 


Sample annual feeder losses before PRISM loss minimization



The same sample feeder showing annual losses after PRISM Loss minimization



Loss Minimization (LM) Brochure PDF


ACS Loss Minimization Product Brochure