How big data can tackle commercial building energy


There’s a tremendous opportunity to reduce energy consumption in commercial buildings, which currently account for more than 40 percent of the nation’s total energy use. According to Pike Research, energy efficiency projects could eventually save $40 billion in annual commercial building energy spending each year. However, at present investment rates, only a small part of that value will be captured.

So why are projects with positive returns going unfunded? Beyond limited access to financing, a large reason is that market participants – energy service providers, utilities, building owners and investors – don’t know which projects to fund. The standard process of manually identifying energy conservation measures is just too slow and expensive, not to mention unsystematic.

But a new breed of data analytics software solutions is aiming to break this gridlock. Using sophisticated algorithms to analyze large data sets for buildings, energy savings opportunities can be rapidly evaluated at low cost, buildings can be prioritized by how much energy they can save, and efficiency recommendations can be delivered – all with minimal human involvement.

At Retroficiency, we focus on this new opportunity. But rather than discuss different vendors and approaches, I’d instead like to focus this article on what data these platforms are actually analyzing. Why? Because by understanding the main classes of energy-related data that can be analyzed, organizations can begin to distinguish between the capabilities of available software platforms and how they can be used at various stages of the process.

Two types of building energy data

Energy efficiency data analytics can be broadly segmented into two classes of information:

  1. Energy interval data
  2. Building asset data

Energy interval data is defined as a record of energy consumption, with readings made at regular intervals throughout the day, every day, over an extended period of time. Depending on the transmission & distribution utility, the sophistication of the interval metering device, and the needs of the user, data may be available in intervals ranging from as short as one minute to as long as an hour.  (Typical is 15 minute increments, and for those of you paying attention, that’s 35,040 data points in a year.)

Interval data can be obtained from utility company meters, smart meters or building control systems.

Building asset data, on the other hand, is information related to general building characteristics and energy systems in the building. This could include basic building information such as the year built or facility use type, to more specifics related to energy consuming equipment, such as heating ventilation and air-conditioning systems, lighting and lighting controls, envelope and windows.

Asset data is available through a variety of sources – including directly from an on-site facility manager, building owner or through energy audits.

Both interval and asset data contain significant hidden insights about a building’s energy efficiency performance and can provide insights about opportunities for improvement.  When this data is combined with sophisticated analytics and correlated to other data sources the underlying information can be turned into a deep understanding of building energy performance and what can be done to curb consumption – without ever visiting the building.

Uncovering insights

By definition, interval and asset data are each more oriented towards certain types of insights. Interval data represents an unbiased view of historical energy consumption. Analyzing changes in daily and seasonal consumption and relating them to weather data, for example, can tell us a significant amount about a building’s operational performance. Are the building’s systems aligned with occupancy hours? Is simultaneous heating and cooling occurring? Are lights being left on at all hours of the night?

Beyond these operational measures, it’s also possible to take interval data and disaggregate this consumption into end use loads – heating, cooling, lighting, etc.  At this point, analytics can be applied to determine whether these systems are consuming an efficient amount of energy or not, and suggest that system retrofits be considered.  However, because interval data does not directly tell you what, for example, specific lighting technologies (T12s versus T8s) are present, further information and calculations are required before determining the specifics of the retrofit suggested.

On the other hand, asset data can include very detailed building system specs. What are the building’s heating and cooling systems? How many lights are in the building and what type are they?  What are the main characteristics of the building roof, walls and floor?  This information can tell us much more about specific retrofit technologies that can be implemented.

Whereas interval data may lead to an insight that ‘your lights appear to be consuming too much energy’, building asset data analytics platforms are able to say, for example, what types of lights are installed in the building (e.g., 40 Watt T-12 Fluorescent Lamps) and determine the related best retrofit (e.g., Super T8’s).  This enables instant calculation of the efficiency gains and cost savings from making a specific building retrofit upgrade.

Combined, interval and asset data can serve as the basis for more effectively reducing commercial energy consumption.  If industry players across the energy, software and real-estate sectors can continue to find ways to drive greater access to this information, data analytics can allow for rapid assessments of building energy performance, savings prioritization, and evaluation of efficiency measures. This software-based approach can go a long way to reducing the current industry bottleneck that exists, unlocking energy reduction opportunities more quickly and more cheaply – ensuring we capture the full value that energy efficient technologies and energy management best practices can deliver.

Bennett Fisher is the CEO and Co-Founder of Retroficiency. Retroficiency enables energy service providers, utilities and building owners to cost-effectively prioritize high potential commercial buildings and evaluate thousands of energy efficiency measures in minutes. For more information, please visit

Image courtesy of RB Boyer, bgreenlee.



If I see the term “big data” on this site one more time, I will officially boycot. The scale of data here is not “big” by IT standards.

james ferguson

Great article and in this case singing to the choir – I am convinced – It is clear that scope exists for enhancement of efficiency of existing building stock.

However while big capital buildings grab headlines for their expensive retrofit and new-builds focus on Zero carbon and LEED ratings there is a huge opportunity in the mass of boring but essential and significant building stock that will never see tens of millions spent, but none-the-less for a very low investment in non-intrusive data-models based on consumption and weather load can see identification of often huge performance anomalies.

Pulling up the lowest common denominator is arguably less “sexy” but simple disciplines like weather compensation, good schedule keeping and appropriate monitoring and dimensioning of plant will make a huge difference over the coming years.

Comments are closed.