I recently presented at the Smart Grid Edge Analytics Workshop held at the Georgia Tech Global Learning Center, sponsored by the National Science Foundation. It is rare to find a conference that brings together academic, industry and research organizations in a small enough venue to openly discuss real, concrete challenges, but this conference achieved it.
A primary focus of the discussion was the intersection of grid analytics and customer loads. Many utilities discussed the opportunities and challenges inherent in using analytics to manage peak loads, like electric vehicle (EV) charging stations, and better balance system load. Many utilities had their own data science teams exploring improved customer segmentation, which enables highly targeted efficiency programs.
It is exciting seeing data analytics and machine learning applied in ways that can directly influence so many lives.
A big theme in many of the conversations was how to respond to the peak load and supply profiles produced by renewables and EV charging. One option is to shift the load to a less expensive time, with something like BuildingIQ’s optimizer tool. New EV charge controls do something similar at a single load level – spreading out a vehicle’s charging over low-peak times.
However, there are other more active options to not only reduce the load but also actively manage it with demand response (DR), and to use buildings’ thermal response and systems as a battery to actively dispatch or absorb energy. Buildings can use their high thermal mass in combination with thermal storage systems, like domestic hot water tanks or ice store systems, to convert energy into thermal energy that can later be released to offset building thermal needs.
A targeted, more granular, demand response strategy is another extension of this approach. In order to understand how analytics changes the game, consider demand response at a macro and micro level where in macro is a set of buildings and micro a single building. Many utilities and companies currently offer DR at a whole-building level, where the whole building is set back to save energy. Utilities have problems with program uptake – meaning not every building that could benefit from DR incentives in enrolled, and some that are enrolled can not drop loads at sufficient levels. Furthermore, determining which buildings to enroll is not easily accomplished. That mean end-users have to make the call, and sometimes that call is not to participate. But what if the utility were to pitch, “Your main student center, Engineering buildings 1, 3, and 4 qualify for a maximum demand incentive of $XXX. Let’s chat.” Anybody who has ever done any sort of marketing or sales knows that specificity and relevance drive attention at dramatically higher rates than generalizations. Understanding a building’s “battery capacity” can make this notion a reality.
At the micro level, it is usually only certain zones or equipment that drive the performance of the system. Changing setpoints in zones that do not directly drive whole building energy introduces unnecessary comfort issues with no energy benefit. At BuildingIQ, we are exploring the use of a new tool that can identify what equipment, setpoints or zones have the highest impact on energy. This allows building owners to focus their strategies on targeting the equipment and areas that matter, like that one critical zone that drives an air handler’s response. The interesting thing about this tool is it can look at any time-series data and find the correlation between any set of points. I wonder how much this tool could be applied to a utilities’ portfolio to look at each building’s meter data and identify what buildings drive the system load profile at any given time? This would allow utilities to target their DR programs to the client and load types driving their system.
Finally, utilities and regulators explored the concept of using digital twins for planning or rate and policy design. By building a digital twin of the entire grid, they can simulate different scenarios and projected load changes to get immediate feedback for critical decisions. The next step is the idea of using a building’s digital twin, like BuildingIQ’s, to then model a building’s response to these policies or how different rate structures would affect a client’s bill. Using a building-level digital twin in this way could uncover what policies have the biggest effect on building system design or occupant comfort, or it could reveal what rates affect certain buildings types. Combining these demand-side digital twins with a grid-level digital twin could produce the holy grail of policy. Something that shows stakeholders the immediate effects of abstract policy changes or rate structures without expensive trials and risk.
Overall, I was incredibly impressed with the research and innovation displayed by all participants at the Smart Grid Edge Analytics Workshop. The presentations demonstrated how grid analytics and grid 3.0 are not just theories but are being implemented today. Now it is up to demand-side providers to find ways to capture these opportunities and create a truly integrated system.
Chris McClurg is Sr. Product Manager, Software & Services at BuildingIQ, and a mechanical engineer focused on energy efficiency in large portfolios and net-zero developments. Chris has worked on deep retrofits, integrated design, integrated project delivery, and buildings as a grid asset. She is a PE, CEM and LEED AP certified.