August 2nd 2011

Often when folks talk about data mining or business intelligence systems for smart buildings, you hear them mention the need to drill down into their data. It’s a given that you’ll have to dive deeper into a mountain of data to find the nuggets of gold (insights) that help you operate your buildings effectively and efficiently. With all the data it’s possible to collect from modern building automation systems, from meters, from sensors and from all kinds of smart devices, it will sometimes seem that you’re looking for a needle in the haystack. 

At the same time, you will also hear people complaining about information overload. They say that they’re deluged with a sea of useless data, which they can’t understand or to act on. You’ll hear folks worry about going down rabbit holes or getting lost in the weeds of excess data.  Certainly you don’t want data integration for your new building systems to turn into an academic science project.  You want results. You want a financial ROI.

These days it’s easy to get overwhelmed by the quantity of information -- not only internal data from your own buildings, but also external data from utilities, benchmarking services, energy markets, service providers, even the weather service.

Luckily there are new software systems available to help you synthesize all of your data wisely while avoiding analysis paralysis. To achieve the right balance for your organization, you will need tools that let you dig deeply into your detailed data while automatically highlighting exception conditions and important facts, so you can spot trends or opportunities to act and save money. More advanced systems can even recognize patterns in your data, embedding sophisticated algorithms like linear programming or neural networks, that can essentially learn what to do about your data and automatically tune your building systems accordingly.

If you manage multiple buildings in a commercial or corporate real estate portfolio, across a campus or in a retail chain, you can conduct internal benchmarking by looking at key performance indicators and comparing that metric across selected individual buildings (or groups of buildings).

Ultimately, this can be a competitive advantage if you can simultaneously manage both the high level big picture and the low level detailed view. A configurable data collection engine, flexible database and advanced charting or visualization tools can let you quickly and seamlessly navigate from detailed data in space and time to a broader 50,000 foot view of all your building data. For example, data about a single KPI from various individually sub-metered floors, work areas or circuits (loads) can be quickly normalized based on time of year, building size, weather, activity (e.g. occupancy, production). This can help you to identify and to begin exploring any exception conditions.

From the same system, with the same user interface, you will be able to zoom out and look across buildings at a macro level, for example monthly energy costs per square foot for all of your buildings based on user-specified selection criteria (e.g. compare high level data for your medium size retail sites in California, or for all buildings that had HVAC retrofits more than one year ago).

The sea of data that you will need to navigate is bound to get wider and deeper as the new energy paradigm unfolds. The cost of detailed data collection continues to come down quickly.  Not only will your buildings get smarter and emit more data, the environment in which they operate will become equally packed with data, with the smart electrical grid, distributed energy generation, plug-in electric vehicles, demand response events, real-time pricing programs and so on.

All this will result in great new opportunities for you to identify the most cost-effective energy efficiency retrofits and automated control strategies for your buildings.

On the growing sea of newly available data, various new reporting, evaluation, measurement and verification requirements are floating to the surface, often driven by regulatory or compliance issues. For example, your company may be adopting a corporate sustainability program or need to comply with regulatory mandates for greater efficiency, such as the Energy Performance of Buildings Directive in Europe. You may need to provide detailed data for your USGBC LEED rating, ENERGY STAR labeling, for your local utility’s energy efficiency programs or to document your savings to earn tax credits.  If you have generation assets, you may have to provide reports also to the US DOE and or the EPA, in addition to state and local government.

Another way you may choose to leverage detailed data in the future would be to create derivatives like white certificates. Already in use by three U.S. states (Connecticut, Pennsylvania, and Nevada) with more expected to follow, these are tradable financial instruments to certify energy consumption was reduced by a specific amount, for example to support legislated energy efficiency resource standards.

Beyond derivatives or regulatory compliance, the true value of operating buildings efficiently of course is simply to save money. Whether or not you’re a “numbers person”, you will need to delve deeper into your building data to find exactly where you can keep saving 5-10% on energy costs each year.

Some people will claim that energy costs as trivial or uncontrollable. Sure, that may be true for some firms. But collectively for the U.S.A., energy amounts to the better part of a trillion dollars each year, or approx 5% to 10% of GDP, depending what you count as energy. And smart devices are making it affordable for you to control what used to be uncontrollable or too costly to automate.

Do you even know much energy costs your own organization? Naturally, that’s the best place for you to start, at a high level (e.g. your total annual energy costs) then later you can do a deeper dive into your buildings (e.g. your electricity use for lighting during the past two hours in your third floor conference room or in the northeast corner of your warehouse).

The Victorian scientist Lord Kelvin (a.k.a. Sir William Thompson, 1824-1907) put it best:

When you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge of it is of a meager and unsatisfactory kind; it may be the beginning of knowledge, but you have scarcely, in your thoughts, advanced it to the stage of science.

To really understand your energy performance, you have to consider granular data to pinpoint your problems in time and space. You need both temporally granular and spatially granular data and control.

By temporally granular, I mean collecting data on a closer to real-time basis – let’s say every minute, quarter hour, hour, shift or day instead of every week, month or year. Another twist on temporal data is about the future, by which I mean production or occupancy schedules as well as weather forecasts and future prices. If I know projected electricity and gas prices, along with the weather forecast and how many people are scheduled to attend a company meeting (for example) in the third floor conference room tomorrow afternoon, you can consider all these factors intelligently in deciding to pre-cool or pre-heat my building structure, or potentially to re-schedule the meeting to another time or place. When something happens you don’t expect, e.g. a weather event or externally driven unplanned event or surprise, such as a surge in orders for your product or services, having the more granular data and control will help you to react in real-time and evolve into more of real-time energy management firm.

By spatially granular, I mean knowing KPI’s and building characteristics for your individual spaces (not just your energy use, but also availability of each space for controls, occupancy sensors, security, smart devices or even maintenance activity scheduling). By continually tuning individual spaces within your smart building based on occupancy or business activity, weather and your real-time energy pricing model, you can optimally implement spatially granular control strategies for lighting and to a lesser extent for HVAC and other systems.

One example of a strategy that leverages both temporally granular and spatially granular data is the strategy of powering off your VoIP phone PC, and POE (power over Ethernet) ports as they are needed by your building occupants, instead of leaving all these powered on all of the time, which is still commonly done in most companies. Side benefits of this include security (it’s hard to hack a phone or PC if there’s no power flowing to it) and potentially lower maintenance costs, especially with advanced solid state digital devices like LED lighting system or tablet computers. On-demand powering such systems on and off becomes more acceptable from the point of view of customer experience – how building tenants or employees experience the buildings. From a maintenance point of view, wear-and-tear costs are not an issue as they are for starting and stopping older fluorescent lighting or disk-based PC technologies.

As an example of granular control, you should be able to easily configure your data collection system to collect data on your cooling system’s chilled water inlet and outlet temperatures, flow rates and electricity demand in kWh down to the hour or every 15 minutes. Too many firms only look at their monthly electricity, gas and water bills from the utility. These are too high level and too long after the fact to help you identify a set of improvements between your HVAC systems, lighting, pumps or plug loads like computers.

To get started, once you have measured which facilities are your biggest energy hogs or most energy-intensive, then you can invest intelligently in granular data collection for just those larger facilities. You may ask your small facilities to simply report their monthly energy bill data, or you could exempt your smallest facilities from reporting until data collection systems come down further in price.

Keep in mind data collection will never be 100% free. You will need to assess the financial ROI before investing in granular data, even before you collect basic monthly energy billing data. It is increasingly possible to automatically gather and share such data, making energy and building management business process outsourcing more feasible than ever before via a third-party BPO firm.

Only you can decide how much detailed energy information you will need to collect. Meanwhile it’s a good bet costs will keep falling each year: installation costs as a result of wireless technology; hardware costs as a result of Moore’s Law; and integration costs as a result of new and evolving standards for data interchange between parties in the landscape.

The bottom line is, the costly devils are in the details. As long as you don’t overdo it, diving deep into your data is the best way to start finding savings in smart buildings. Choose a building software system that helps you identify outliers and exceptions, and that also can automate responses by talking back to your building automation and control systems.