Optimize Now
Has your computer or mobile device ever rebelled at your commands? Acted slow, not responded at all, or rebooted on its own? Maybe the better question is – what have you done after experiencing such misbehaviour? Chances are you have come across optimization software promising to fix all your computer headaches and getting it to work better than ever. You may have even clicked that Optimize Now button on your handheld device to awake built-in utilities and perform maintenance tasks that clean up files to free up memory, check for viruses to isolate them, and shut down runaway apps to extend its battery life. This concept of optimization also applies to buildings. Wouldn’t it be great to have that Optimize Now button at the controls console inside the operations room? In effect, the task to optimize building performance is the main objective during the occupancy and operations phase of new construction commissioning. While not a single button – and certainly not automated – this process entails a series of tasks to identify opportunities to fine-tune schedules, sequences, and set points in accordance with actual building occupancy and use. Once commissioned, however, the reality is that buildings will not function at that high performance in perpetuity as they initially did. Throughout the life of a facility, physical spaces and occupancy requirements change, equipment and devices fail, systems are added, operational sequences are altered – all these driving the need for building systems optimization. Ongoing Commissioning (OCx) [1] is essentially the response to the need for persistent optimization to keep buildings performing at their best, always. More on this later in this article.
What is Automated System Optimization?
-
The technology that enables the continuous analysis and automatic adjustment of building automation system parameters to optimize the energy utilization and process output – including occupant comfort and emissions reduction – from systems operating within buildings.
Optimization, or mathematical programming, refers to the “collection of mathematical principles and methods used for solving quantitative problems” in various disciplines.[2] While control theory applied to physical processes and systems has close ties with optimization techniques, this blog will keep the discussion at a high level. So what does it mean to optimize? Here is a simple definition: “to make [the object] as effective, perfect, or useful as possible”.
In the Primer on Organizational Use of Energy Management and Information Systems (EMIS) [3], a Better Buildings publication authored by the Lawrence Berkeley National Laboratory (LBNL), Automated System Optimization (ASO) is described as a system-level EMIS tool used to “dynamically modify Building Automation System (BAS) control settings to optimize HVAC system energy usage while maintaining occupant comfort.” According to this resource, “two-way communication with the BAS is the distinguishing feature of ASO solutions”, adding that, “these tools both read data from the BAS and write analytically-based optimal set points back to the BAS, based on measured indoor, outdoor, and energy price conditions.”
Can’t the BAS Do It?
Basic system-level optimization, yes. Likely, your BAS is already performing basic automated optimization. For instance, reset strategies, where temperature and pressure set points in primary systems such as air handlers, chillers, and boilers are adjusted regularly based on demand from their respective secondary systems, have been standard implementation for decades. Optimum start and stop algorithms are examples of system optimization whereby the BAS attempts to reduce energy consumption by finding the optimum start and stop times required to maintain comfort levels during occupied hours. Given that ASO functionality is not designed to replace a BAS, basic optimization should always reside within the building automation system. In general, no dependency should be created between the BAS and the ASO, especially when continuous adjustment and real-time controls are required.
Advanced building and system-level optimization, however, is more complex and requires a more comprehensive view of the past (i.e., historical perspective) to forecast the future (i.e., predictive analysis), capability seldom found even in modern BAS solutions today. On the other hand, supervisory EMIS platforms have the data infrastructure and can be made aware of external influences coming from utility providers, current and forecasted weather, and building usage and occupancy, providing the visibility needed to perform more advanced system optimization strategies. Moreover, when disparate multi-vendor BAS platforms exist in a building or portfolio, EMIS solutions unify and simplify the integration and deployment of ASO measures at the top layer.
Advanced Optimization to Curb Demand
Perhaps the earliest system optimization requirements that BAS platforms were tasked with came from the need to curb electricity demand. Control integrators would implement load shedding so building owners could avoid the large financial penalties associated with peak demand or time-of-use tariffs imposed by utility providers. For this, awkward and often flaky integrations were required to get the BAS to trigger custom control sequences, adjust schedules, and command global set points to shut down equipment or delay their start-up, reducing or shifting power needs as a result.
Today’s Demand Side Management (DSM) strategies call for more advanced predictive analytics to model and forecast load profiles ahead of time, functionality that is rarely possible at the BAS. Furthermore, the Grid-interactive Efficient Buildings (GEB) Initiative [4] aims to “remake buildings into clean and flexible energy resources by combining energy efficiency and demand flexibility with smart technologies…” using building analytics to access data outside the physical boundaries of buildings, and the computing power within the confines of accessible and secure cloud infrastructures.
Why Use Analytics Software for Automated System Optimization?
-
Because system optimization is a data-driven process – analytics software captures historical data to assess past performance while a continuous flow of new data monitors system behaviour looking for optimization opportunities.
-
Because it is better positioned to understand the various factors that influence performance – analytics software has greater visibility in how a building is used and a broader understanding of the systems interacting within it.
A Data-driven Process
The EMIS toolbox includes tools that provide the monitoring and analytic capabilities to support data-driven continuous processes such as Monitoring-Based Commissioning (MBCx)[5] – defined by LBNL as “a way to maintain and continuously improve building performance over time”. MBCx methodologies are often incorporated in OCx processes where commissioning service providers collaborate with building owners, operators, and others to improve and maintain building performance.
The closely-knit relationship between optimization and performance is evident when one reads how the sustaining phase – the third and final phase of the OCx process – is described in the Ongoing Commissioning Best Practices released in 2020 by the Building Commissioning Association. [1] Considering that “the primary goal of a successful OCx project is the ongoing, sustained building performance for the long term, and ideally for the life of the building, [OCx] provides the building owner with assurance that the investment on building systems and the optimization of those systems continues to operate in an optimal manner.” Not only does it mention the need to optimize systems, but also that the process of optimization needs to be done optimally!
Integrated for High Visibility
Most EMIS platforms are designed to integrate data from multiple sources, including but not limited to BAS networks, IoT devices (Internet of Things), energy metering systems, building occupancy monitoring, weather data providers, and utility data platforms. This integration can occur directly to self-contained, siloed systems, or via centralized data repositories such as data lakes and data warehouses. Bi-directional integration is needed when building operations and maintenance platforms, as well as asset management systems, are designated consumers of the raw and derivative data. In short, the integrative nature of supervisory platforms provides the visibility that is seldom possible from the individual systems operating within a building. And this is indispensable to the leveraging of all possible sources of information to make more informed decisions on the use of resources – whether natural, human, or economic – to optimize how systems operate and interact with each other.
Now let us get more practical. If you already have a system-level EMIS in your facility, the following steps should provide some insights on how to begin implementing an ASO strategy. However, if your organization is currently considering implementing EMIS technologies, a couple of guides, one on evaluating Fault Detection and Diagnostics (FDD), and a second one on evaluating Energy Information Systems (EIS) might be more appropriate here.
How to Expand the Use of Analytics Software with Automated System Optimization?
-
Establish Optimization Objectives beginning with opportunities that are easier to implement. Target specific systems, and define candidate set points that would benefit from optimization functionality.
-
Understand the 2-way Interaction between the EMIS platform and the Building Automation System. This includes connectivity requirements, capabilities when a 2-way interface is established, and all features around safeguards.
-
Integrate Optimization Functionality in existing operational workflows. Establish procedures to guide building operations prior to initiating optimization measures from the EMIS platform.
The first step is to establish optimization objectives. Following the completion of building commissioning, all systems are expected to operate optimally. Ideal schedules, control parameters, and set points would be in place. EMIS platforms able to capture these values regularly can let operators know when, for whatever reason, they are changed. If the change is good, operators would save the new values; if the change is not good, operators would revert to the old values. This last action would be enhanced if users could assign the required change to an automatic function that would carry out the value reversion. In other words, here is where ASO functionality could take over and automatically update the schedule or set point as needed. Figure 2 below shows the “Latest Value” and the “Golden Standard” or “baseline” value side by side, and recorded under “Action History”, a successful action by the ASO reverting the value from 72°F to 68°F on March 11th. Can you think of other control parameters and set points that could be auto-corrected in a similar fashion?
Figure 2. Automated System Optimization functionality in action reverting a current value to its baseline.
Figure 3. Cloud-based FDD platforms with different 2-way communication approaches to achieve ASO functionality. (Source: Can We Fix It Automatically? Development of Fault Auto-Correction Algorithms for HVAC and Lighting Systems [6])
Lastly, the task to integrate optimization functionality in existing workflows is vital as facility personnel begin to interact with new automated functionality to understand its benefits while reducing manual tasks. As their level of confidence increases, more complex automated system optimization functions could be deployed.
To emphasize once more, the implementation of ASO solutions should not involve the removal or relocation of the various control programs and possibly thousands of lines of code as these are still required to reside within the BAS. Now, given that ASO aims to improve system performance, this is probably an excellent time to review those control programs and ascertain that they are following the designed specifications for each system as documented in the O&M manuals… so time to dust off those artifacts!
ASO is one tool in the toolbox known as EMIS. Others include Energy Information Systems and Fault Detection and Diagnostics. To learn more about these tools, please contact us, and visit our fundamental series blogs giving answers to What is an Energy Information System? and What is Fault Detection and Diagnostics?
End Notes
[1] Building Commissioning Association. 2020. Ongoing Building Commissioning Best Practices
[2] Britannica. 2021. Optimization
[3] Lawrence Berkeley National Laboratory. 2015. A Primer on Organizational Use of Energy Management and Information Systems (EMIS). U.S. Department of Energy Better Buildings Initiative.
[4] Energy.gov. 2021. Grid-interactive Efficient Buildings. Office of Energy Efficiency & Renewable Energy.
[5] Lawrence Berkeley National Laboratory. 2018. What you need to know about MBCx
[6] Lawrence Berkeley National Laboratory. 2020. Can We Fix It Automatically? Development of Fault Auto-Correction Algorithms for HVAC and Lighting Systems. Energy Technologies Area. p.3-159
Written By:
CopperTree Analytics
More By This Author
