Managing Thermal Systems with Axiom Cloud's IoT Platform
Along the journey of deploying and operating Refrigeration Batteries (RB’s) at supermarkets in California, we invested heavily in our software and controls to maximize the value that our thermal storage assets deliver. As we optimized the Refrigeration Battery and associated refrigeration systems, however, we discovered that our software alone could unlock meaningful flexible capacity that is already inherent to low temperature refrigeration systems. Specifically, by pre-cooling low-temperature cases and then letting the temperatures drift back to their initial setpoints, we could enable facilities to participate in energy bill management and grid services - without additional thermal energy storage hardware. By intelligently modulating these refrigeration systems, we would be able to turn refrigeration systems into "Virtual Batteries" that could be deployed quickly and efficiently.
At this point, we decided to give our software stack a name - Axiom Cloud's IoT Platform - and continue to use it to operate our Refrigeration Battery while also offering it as a stand-alone active power management solution. The main realization here was that this platform is an ideal solution for thermal systems with some inherent, currently underutilized flexibility, such as freezer sections of supermarkets. The Refrigeration Battery, on the other hand, is an ideal energy storage solution for systems that do not have much existing flexibility, such as produce sections of supermarkets which must be maintained at very precise temperatures in order to ensure food safety. By deploying the platform not only onto RB’s, but also onto existing thermal systems with more flexible setpoints, we are able to reduce utility bills, introduce new revenue streams, reduce maintenance spend, and decrease equipment failure at a much wider range of facilities than ever before.
How It Works
Axiom’s IoT Platform is a lightweight, software-based addition to any refrigeration system that transforms the entire building into an intelligent, flexible, optimized asset. Upon being deployed at a customer site, the first step is to build a “digital twin” of the building’s main electricity meter and thermal systems; both of these models are improved over time, as more data is collected, using machine learning algorithms. After this, the AI engine uses forward-looking data streams such as weather predictions, load predictions, and market predictions to generate forecasts for both building power and thermal system operation.
These models and forecasts, which are continuously updated and improved as time goes on, are what we use to inform the Artificial Intelligence algorithms that work to optimize the building’s performance. These algorithms autonomously unlock the existing flexibility of thermal systems by predictively pre-cooling refrigerated spaces, followed by intelligent load-shedding of compressors and condensers. Using these methods, the platform is able to generate up to 40% of the savings generated by the RB alone, all without additional hardware. In addition, by further leveraging AI, the platform is able to help facility managers and technicians solve their biggest maintenance problems by predicting compressor/condenser/evaporator failures and refrigerant leaks before they occur, thus allowing them to plan repairs ahead of time rather than simply reacting to issues as they arise in real time.
As an illustration of the demand management capabilities of the Platform, the following graph shows how the AI engine chose to operate over 24 hours at one site in California.
Legend:
Baseline Building Load (kWe) - Blue
Actual Building Load (kWe) - Green
Outdoor Ambient Temperature (oF) - Red (top graph)
LT Refrigeration System Power (kWe) - Purple
Rate of Charge/Discharge (kWe) - Yellow
LT System SST (oF) - Red (bottom graph)
As you can see, without any additions or changes to existing hardware, the platform was able to reduce the building’s maximum demand on this day from 516 kW to 454 kW, a reduction of 62 kW - all autonomously and without allowing product temperature to go above the existing setpoint of -18oF.
By operating intelligently year round at fleets of sites, Axiom Cloud's IoT Platform is able to provide customers value in many different ways. In the next section, we’ll go into those value streams in further detail.
Current Value Streams
Through intelligent pre-cooling and load-shedding of thermal systems, the IoT Platform is able to provide services across fleets of buildings that are similar to those provided by a traditional battery. In fact, the services provided are so similar to those provided by lithium-ion battery, that we call our current suite of services a “Virtual Battery." These services currently include:
Energy Bill Management
Virtual Battery can reduce the energy costs of a refrigeration system by 10-20% by actively managing demand charges and shifting energy consumption away from expensive peak periods. After using an advanced optimization that takes into account thermal constraints, electrical constraints, weather forecasts, utility tariffs, and more, Virtual Battery modulates thermal systems to reshape the building’s load profile and reduce costs as much as possible.
Grid Services
Virtual Battery also allows customers to unlock a new revenue stream by transforming building portfolios into Virtual Power Plants (VPPs), thereby allowing them to participate in grid services programs such as demand response and capacity markets. By turning fleets of refrigeration systems into VPPs, Virtual Battery monetizes the flexibility of these systems in a way that has never been done before.
What’s Next for Our Software Solution?
We have been hard at work developing features that will increase the value of the IoT Platform and make the operation of refrigeration systems as simple and efficient as possible. Once development is complete, these features will reduce the costs and headaches associated with the operation and maintenance of refrigeration systems by predicting failures before they occur and improving system performance over time. Currently planned services include:
Continuous Commissioning
The platform will be able to reduce electricity bills by an additional 10%, increase average Coefficients of Performance (CoPs), and ensure persistence of savings by autonomously monitoring and improving systems over time. Floating condensing temperatures, floating suction temperatures, and anti-sweat heater controls often “drift” shortly after they are implemented, and continuous commissioning ensures these benefits are sustained over time.
Predictive Maintenance
The platform will reduce failures of mission-critical refrigeration and reduce overall maintenance costs by autonomously predicting compressor/condenser/evaporator failures and refrigeration leaks before they occur. This will allow site managers and technicians to look ahead and plan for repairs instead of simply reacting once they’ve already occurred.
We are very excited to be sharing more details about Axiom Cloud’s IoT Platform, our software-only solution for buildings with large thermal systems. If you’re interested in learning how to optimize your grocery store or refrigerated warehouse , contact us today.