Managing a multi-player community of renewable consumers and producers successfully
Utility and public interest in renewable integration within distribution networks has rapidly risen in popularity. For the public, general sentiments concerning global warming and the environment have been further stoked by subsidies and economic incentives. Utilities face the technical challenge of anticipating the impact of ever increasing amounts of both utility-owned and private sector-owned renewable generation, while keeping their power grids safe, reliable, and cost effective.
Preceding the interest in renewables, the initial advancement of the smart grid that emerged with the greatest notoriety was automatic switching functionality, or self-healing, which improved network and individual feeder reliability. Self-healing technology, often called FDIR or FLISR, is the idea of providing tools (automation, software, network modelling) which allows the network to figure out what is wrong and to automatically execute a solution which restores power in real-time. Utilities that adopted these early technologies gained the foundational understanding to use these sophisticated applications to augment and support their manned operational control centers. Those utilities will notably benefit by having laid the automation foundation that is critical to successful management of renewable deployment. The technological progression from self-healing to power quality to Distributed Generation (DG) management is a natural evolution of increasing complexity and as the degree of renewable generation deployments increase, the reliance on these advanced tools to support ongoing safe, reliable, cost-effective operations becomes mandatory.
The greatest challenge regarding renewable deployment is that the amount and variability of generation, along with the various points throughout the grid where the generation is operating (injection sites), is always changing. The power grid was not originally designed to operate in this manner, so in almost every instance today, impact of significant renewable injection within a feeder is unknown and unmeasured. The good news is: the original principles behind transforming our old power delivery networks into Smart Grids was based on the idea of using technology to monitor everything happening with the flow of electricity, and using mathematics and physics to make real-time decisions on how the grid should operate. The complexity of developing the network model and load flow solution algorithms is at the heart of the Smart Grid. Those same tools allow utilities to accept and manage the continuously changing flow of diverse renewable generation sites without risking the safety or reliability of the underlying power delivery network.
Changes in Automation
Smart Grids evolved first with the deployment in the field of automated switches and re-closers, the tools that utilities use to re-route power flows. Next, many of those same utilities took the further step of adding software in the form of Distribution Management Systems (DMS) and advanced algorithm-based applications which allow utility grid operators to assess in real-time what is going on in the network and to take action (remotely operate equipment in the field) to maintain the best and most reliable operation. In an area where new renewable generation sites are being added, these smart grid technologies that were installed to enhance reliability and power quality in the feeder’s operation will initially operate unimpeded with minimal deployment of distributed generation. At low levels of renewables deployment, the application of distributed generation within the feeder resembles a negative load (because instead of consuming electricity, that site is adding net electricity to the grid). In fact, many DMS solutions simplify the network calculations and modeling of injection points within the feeder in this way.
As more and more renewable generation is brought on-line, the over simplification of treating the distributed generation as a negative load is ineffective. Significant injection will soon threaten the feeder’s stability without a feeder load flow analysis that considers the dynamic nature of the injection points. The load flow itself must be capable of handling multiple sources.
To deal with this new challenge, smart grid technologies specifically Voltage Control (such as Integrated Volt/VAR Control, also known as IVVC) must be retooled to include control of the inverters at each renewable site with the objective of increasing the coordination of each feeder’s maximum injection capability. Switch plan optimization applications must likewise establish an objective function, which maximizes the feeder’s ability to accept the maximum injection safely. The solution to successful deployment of significant distributed (renewable) generation injection, builds directly on the classical smart grid automation applications.
The architecture of systems which support distributed generators, distributed energy resources, renewable deployment or energy storage systems must accomplish the following primary objectives:
- Preserve the viability of the feeder’s operation at all times.
- Maximize the availability of the DGs/renewable generation, meeting a user-selected business case.
- Maximize the deployment of energy storage assets, meeting a user-selected business case.
A modular renewable management architecture that is well-suited to accomplish these objectives is performed by two main systems: the DMS and Distributed Energy Resource Management System (DERMS). Since the two primary objectives of a renewable network may work in opposition to one another, an architecture that is dedicated to maximizing the two objectives as their primary mission is required.
The operation is best served if the division of responsibilities of the systems ensure the following:
- The DMS is eminently positioned to perform the secure, reliable and optimum operation at all assets in the feeder under all operational conditions.
- The DERMS is dedicated to forecast and maximize the output and availability of the DG / DER resources under the objective established by the use case selected by the operator.
Other architectures may apportion the responsibilities differently, for example they bundle IVVC as a DERMS responsibility. More important than the architecture, or the assigned mission parameters of the DMS/DERMS, is that IVVC or IVVC/r must always be run as a mission critical automation application. IVVC, like FDIR/FLISR, must be protected by high availability with real-time performance, especially under severely stressed network conditions.
This goal is a challenge with some ADMS solutions that have functionally evolved from a non-`real-time origin. In those cases, a distributed automation architecture such as DERMS is preferred in order to meet the mission critical requirements. Some very large models are equally challenged with respect to maintenance where a smaller distributed model is more easily managed within a distributed network of DERMS processors. In these cases, the DMS is a collector of the distributed automation islands for operator oversight purposes, but it is not placed at risk by a low availability centralized system.
In contrast, if the mission critical requirements can be met along with accurate and timely model maintenance, the DMS centric feeder automation architecture that supports IVVC in the DMS, simplifies the analysis and modeling which is always an important consideration, since the modeling of the distribution network feeder is a continuous and arduous effort. In this architecture, the network model is consumed only in the DMS and it is not necessary to distribute it to the DERMS. In short, the architecture can be a combination of central and distributed intelligence, flexible enough to meet the capabilities and requirements of the mission critical functions.
Modules for Renewables
Not all renewables, or distributed generation, are equal. Therefore, the optimum strategy for their deployment is tailored to the application. The ability for the feeder to be able to safely accept the maximum output capability of renewables such as photovoltaics (PV), without resorting to curtailment, is important. Maximizing the potential can only be accomplished if the forecasted capability of the DG is predicted far enough in advance for the DMS to analyze the hourly impact of the scheduled injection.
If the forecast predicts that the schedule will result in time periods that will exceed the network’s stability limits, there are two situations that must be addressed:
- Can the normal functioning of the Volt/VAR control avoid the violation? If it can, no further action is required. The DMS will handle it.
- If the DMS cannot compensate for network weakness imposed by the injection a more advanced analysis of the options must be made.
In the latter situation, the DMS has the ability to calculate any network configuration and topology that will support the maximum forecasted schedule or the adjusted forecasted schedule. This may result in switching changes, similar to an enhanced self-healing solution, which will transfer loads and DGs from one feeder to another in order to meet the forecast. An appropriate lead-time to affect change is important since typically few feeder-switching devices are remotely controlled. An optimum integration of sending switch plans from the DMS to crew mobile can improve the efficiency of the switching response.