Article Post

Automatic Load Shedding in Active Distribution Networks

Automatic Load Shedding in Active Distribution Networks

Nokhum Markushevich

The Under-Frequency and the Under-Voltage Load Shedding schemes – (UFLS and UVLS respectively) are the dominant load-shedding remedial action schemes in the power systems.

The discussion below addresses mostly the UFLS. However, most of these discussions also relate to the   UVLS.

Currently, when the frequency drops in a bulk portion of a power system, the Under-Frequency Load Shedding (UFLS) schemes disconnect a number of pre-selected distribution feeders following pre-selected settings for pre-selected groups of the UFLS. The pre-selection is based on the notion that there is always load demand on the feeder, and disconnecting the feeder will always reduce the load and raise the frequency.

In active distribution networks, which mean high penetration of Distributed Energy Resources (DER) including energy storage systems with frequency/voltage sensitivities, microgrids,  demand response, etc., it is possible that at some times some feeders will be “generation-rich”. In this case, the UFLS should not disconnect the feeder. At the same time, there may be microgrids connected to this feeder that can contribute to the mitigation of the bulk system emergency by disconnecting from the grid, if they consume power from the grid, or by disconnecting a portion of their load, if they have their own UFLS schemes.

Even when the feeder is “load-rich”, shedding first the microgrid loads by internal UFLS may convert the feeder into a “generation-rich” one. On the other hand, if the UFLS schemes do not work before the DER protection (ride-through) works, the system loses a portion of generation, and the emergency worsens leading to a cascading development.  

Therefore, in the Smart Grid environment, the UFLS will need to be more portioned. In addition to the feeder heads, the UFLS schemes may be located at the sectionalizing breakers/reclosers in the middle of the feeder, at the feeder branches, at the micro-grid PCCs, and at the load switches inside the micro-grids.  This allocation will depend on the placements of the DERs and micro-grids along the feeders.

All these protection schemes need coordination under different pre- and post-shedding conditions and events.

Each of the above-mentioned location may be in one of the two pre-shedding conditions: either “load-rich”, or “generation-rich”. In addition, the total load connected to the UFLS in the microgrids can be either greater or smaller than the net load of the feeder. The load connected to an individual UFLS of a microgrid can also be either greater or smaller than the net load at the microgrid’s PCC. The reaction of the energy storage systems to the change of the frequency should be taken into account, when determining the load-generation relationships. All these multiple conditions are changing in near-real time. Hence, the optimal coordination of the load-shedding schemes in different locations of the distribution grid is also changing. Adding the possible settings of the DER ride-trough functions to these situations increases the number of possible combinations even more.

Hence, a consideration should be given to the capability of a near-real-time coordination of the load-shedding and generation-shedding schemes in different places of the active distribution networks.

In some cases, such coordination can be accomplished with minimum near-real-time communications. For instance, if a microgrid is “generation-rich’, its UFLS can be either disabled, or setup according to the contractual agreements with the EPS. If the microgrid is “load-rich’, and the load connected to the UFLS is smaller than the net load at the  PCC, the microgrid’s EMS should make the UFLS within the microgrid to work first, before the UFLS at the PCC works, trying to keep the connection to the grid. In the opposite case, the UFLS at the PCC should work first to minimize the load shedding in the microgrid. These kinds of solutions can be made by the microgrid’s EMS based on local microgrid’s information.  

If the microgrid’s EMS transmits to the EPS the information about the load/generation shedding schemes setups, the DMS can chose the setups of such schemes at other feeder locations [1].

Conclusions.

  1. Conventional setups of the load-shedding remedial action schemes in active distribution networks may lead to cascading emergencies and load over-shedding.
  2. Local load-shedding schemes within microgrids needed for load balancing in island modes may also operate in connected mode in cases of EPS emergencies.
  3. Placements of the load-shedding schemes in EPS circuits may be extended to the portions of distribution feeders.
  4. The operations of the different load/generation shedding schemes and frequency/voltage control schemes in different places of the active distribution network should be coordinated in a near-real-time fashion.
  5. In many cases such coordination can be accomplished with minimum near-real-time communications
  6. The energy management systems of the transmission and distribution domains, as well as the EMS of microgrids should be upgraded to accommodate the coordination of the load/generation shedding schemes.

Reference.

Nokhum Markushevich, Information Exchange between Advanced Microgrids and Electric Power Systems. Available: http://www.publishresearch.com/publication/2220

Discussions

No discussions yet. Start a discussion below.