Why big data and building analytics aren’t going anywhere: Part 1

January 30, 2019

Posted by: Anasia D'mello

In 2014, businesses around the globe collectively produced a whopping 8.4 zettabytes (or 8.4 trillion gigabytes) of digital content, up from 2.7 zettabytes in 2012. That’s a lot of information to parse — hence the term “big data,” which describes the trend of processing high-volume and highly variable information toform modern insights and optimise outcomes.

Interpreting this mountain of big data for the purposes of constructing or renovating better commercial and residential buildings requires new modes of thinking, such as building analytics.

Building analytics are part of the growing trend toward fuller automation of building systems, a market expected to reach $100 billion (€87.5 billion) in the next four years. These systems have been integrated into larger schemes for sustainable building, compliance, and resource management. HVAC functions play a significant role in how these analytics clear a path for more efficient systems usage and optimised process management, says Kevin Burns is the president of Bob Jenson Air Conditioning.

Inefficiency costs big bucks

Heating and cooling systems consume somewhere between 25-30% of the annual total energy use in residential buildings and between 40-60% in commercial buildings. For instance, a chiller plant consumes about a third of all HVAC-related energy requirements (or roughly one-fifth of a building’s total energy requirements).

Using meter data to identify deficiencies can be challenging, as energy waste is often incremental (and/or non-linear) and will become hidden among a larger backdrop of wastage statistics. Instead, those eye-catching and costly numbers can be cut by nearly half with proper system maintenance and data analytics.

Big data can create thousands of gigabytes’ worth of information on residential and commercial HVAC systems to make large-scale records, which could find historical trends, analyse cause-and-effect patterns, benchmark HVAC performance, and compute any other number of real cost-efficiency metrics.

Kevin Burns

This type of management could reduce normal operation energy costs each year by up to 20% as a building reaches its optimal efficiency, and it can reduce downtime costs by 35-45%. The average return on investment for analytic solutions is more than $13 (€11.3) for every dollar spent.

Visibility at the top

The design of many buildings and their HVAC systems are often inherently inefficient due to different components (e.g. coils, fans, valves) that are not modeled separately owing to coupled dynamics. The sheer number of set points, levels, and feedbacks in any ventilation system makes top-down visibility an uppermost priority.

The latest in algorithmic learning

Building analytics, through Machine Learning algorithms, have the ability to reduce energy demands, account for shifting weather conditions, detect occupancy and comfort patterns, and handle peak load distribution of HVAC systems.

The very latest machine learning algorithms, known as deep neural networks (DNNs), use Artificial Intelligence to solve complex problems by consuming raw data and processing it through many transformative layers. DNNs have only recently been implemented to solve HVAC issues, but early returns suggested a potential energy savings of 30%.

The author of this blog is Kevin Burns is the president of Bob Jenson Air Conditioning 

About the author

Kevin Burns is the president of Bob Jenson Air Conditioning in San Diego with over 29 years of experience in the HVAC Field. He has worked in every aspect of the industry and has trained dozens of people. He has a passion for doing what’s right for each home and customer and sets this standard for his entire team

Comment on this article below or via Twitter @IoTGN