Energy management systems (EMS) both enhance and complicate microgrid design. Schneider Electric’s Andy Haun explains what it takes to ensure that modeling reflects outcome in today’s sophisticated microgrids.
There are countless variables to consider when making an investment into a microgrid, from the electrical distribution hardware, to the control solution and to the types and sizes of distributed energy resources (DERs) the operator plans to use. Unfortunately, when assembled as a system, the microgrid operator often experiences different outcomes than expected during design, which is most commonly attributed to the variable nature of DERs.
While the resilience function in microgrids and its relation to power system controls is well understood, today’s microgrids add energy management systems (EMS) which introduce unique algorithms based on artificial intelligence and machine learning engines that can increase DER functionality and optimize their energy contribution. This recently expanded element of energy optimization is complicated and although intended to increase the value of DERs in a microgrid, can create even greater variance between expected and actual outcomes.
Selecting and sizing DERs for optimized energy provision requires complex analytics of many cooperating variables related to DER costs (capex and opex), applicable utility energy tariffs, net metering requirements, site demand data and geographical constraints. There exist numerous modeling tools that address the simplest and most mature of these combined sources such as solar photovoltaic alone or coupled with energy storage. However, when we add the benefits of combined heat and power (CHP) or baseload generators, the way the control system chooses which assets are best to deliver power to the loads significantly impacts the economic and sustainability performance of the microgrid.
To avoid disparity between modeled and actual outcomes, it’s imperative that microgrid designers understand and correctly model the variables within the EMS they intend to use in the deployed microgrid.
The new microgrid proposition – energy optimization
Using DERs in microgrids to provide emergency power has been well established for decades. Modern power system control technologies implementing resilience are extremely flexible, meaning they can integrate multiple local and utility resources to deliver better and longer duration resilience. System architectures focused on resilience have been widely deployed in critical organizations such as hospitals and data centers for which secure power is paramount.
The new microgrid use case not only embraces resilience and power system management, but also energy optimization. Brought on by less expensive and renewable DERs along with digitization via IoT, energy optimization enables better use of more sustainable sources and lowers costs by making the grid more functional at the behind-the-meter location. Energy management systems orchestrate the optimal use of such DERs.
Understanding the differences between how the modeling algorithms are assuming use of the energy assets versus how the EMS operates them will help microgrid owners design for the right outcomes of their deployed system.
Additionally, an EMS enables the microgrid to take advantage of site behavior, such as how it naturally consumes energy and link site managers choices about the optimal utilization with automated decisions regarding when to run on-site DERs. For example, it manages the choice between buying energy from the grid, generating it locally, storing it locally, and using energy at the right moments according to provisions of the grid, storage and local production availabilities.
Algorithms, models and simulations – oh my!
DER selections are made based on behavioral expectations of a variety of different control system algorithms established by DER manufacturers and energy management system companies to deliver a good sustainability or economic outcome. The way these EMS algorithms are implemented determines the performance of the microgrid’s energy system. When you’re making choices about the size, type and economic value of the DERs, it’s based on some construct of how the control system(s) will execute their behavior.
Now, here’s where things get tricky. Often, the modeling software algorithms used to help people make DER type and size decisions are different from those in the actual energy management control system that will run the system once deployed. If the algorithms for energy management in your design model are not the same as those in the as-installed system, you’ll risk getting very different outcomes. In other words, your modeled system performance, by definition, is not reflected in the as-built operating system’s behavior, thereby yielding unexpected results.
In this situation, the biggest risk for the microgrid developers and owners is economic. If the microgrid does not deliver the expected return on investment, investors or capital allocation leaders will have less confidence in the microgrid or DER, reducing its expected value. Accurately modelling the behavior of the generation assets under control is critical to maximizing the financial benefits and investment justification of the DERs in a microgrid.
Understanding the differences between how the modeling algorithms are assuming use of the energy assets versus how the EMS operates them will help microgrid owners design for the right outcomes of their deployed system. In the best scenario, microgrid designers should use a modeling engine that uses the algorithms in simulation as those for the deployed microgrid to ensure performance outcomes meet expectations.
There will always be some variabilities in modeling outcomes for microgrids with DERs. But, as microgrid use cases continue to evolve including advanced energy optimization, simulation of energy flexibilities must mirror the actual algorithms that will implement them. This is becoming an important conversation with existing and prospective microgrid owners looking for designs that ensure power management, resilience and energy optimization with DERs will meet their financial and sustainability objectives. As consideration between design modeling and actual deployed algorithms expands, it will be interesting to see how the industry responds.
Andy Haun is senior vice president and chief technology officer, microgrids, at Schneider Electric.
