{Encycle} HVAC Visibility Insights Help to Lower Costs, Improve Sustainability

HVAC units are not only among the largest electricity consuming loads in buildings – typically about 40% – but they also represent the loads that are most likely to suffer from both scheduling inefficiencies and mechanical faults. It is therefore important to facility operators, energy analysts and finance personnel alike that HVAC loads are operating as efficiently and effectively as possible to ensure optimal cooling with the least cost and environmental impact.

One might think that monitoring occupant complaints would be a sufficient method to know if something is amiss with HVAC loads; however, this assumption is incorrect for a number of reasons:

If HVAC loads are operating well before or after occupants are at work in different zones of a building, they would never realize that such waste is occurring.

In large, open areas, if one of the HVAC units is functioning poorly, its neighbors may “pick up the slack” by working more aggressively to compensate for the faulty unit. The other HVAC units then must work more than would be normally required to maintain the setpoint temperature. These and other HVAC faults typically only become apparent under extreme heat conditions, such as during summer heat waves when HVAC technicians are already at their busiest, dealing with a flood of emergency repair requests.

Occupants may not always communicate faults to the appropriate facilities personnel, leaving faults unknown until the next round of preventive maintenance occurs. By then, energy and money have been wasted, and further damage may have occurred.

Finally, there is a quandary that facility managers often encounter, where two people standing side by side have two different opinions about whether they’re comfortable or not. Consider a theater employee working at a desk compared to one who just finished vacuuming hallways for an hour. Who is more likely to complain that it’s too hot? So, you may be able to please all of the people some of the time, and some of the people all of the time, but never all of the people all of the time. For this conundrum, we have no solution, but greater insight into HVAC operation can – and will – create an opportunity to achieve the best possible results.

The Devil is in the Details

Valuable insights can be gleaned by understanding exactly what certain loads are doing under different circumstances. These insights are gained by determining when the loads are – or are not – using electricity, compared to what is expected. This applies to both the entire class of loads, as well as to individual loads.

Relying solely on utility meter data is insufficient in achieving the insights needed because it provides data only about the building as a whole, not to mention the fact that just obtaining the data can be an awkward procedure. And while some utilities are embracing the Green Button data initiative, adoption is still rare across the U.S. and Canada, and even then, it cannot provide load-specific data. The optimal method of monitoring HVAC loads is to meter each one, known variously as disaggregation, disambiguation or simply sub-metering.

Sub-metering + demand management = Synergy That Works

Historically, solutions that provide only sub-metering functionality tend to be more expensive than the value they generate simply in terms of scheduling and maintenance efficiencies. However, marrying a sub-metering solution with one that also provides scheduling and demand management/demand response functionality is an ideal way to bundle the benefits into a cost-effective package.

Some energy analysis systems try to infer what specific loads are doing without directly metering these loads. Such systems attempt to discern some form of “signature” in the overall stream of meter data in an attempt to identify specific appliances, although such systems are unable to distinguish multiple instances of the same load (e.g., three HVAC units of the same make and model at one location). Thus, these types of systems have only been applied to residential settings rather than commercial buildings, where this inferential approach has the possibility of discerning different loads. However, even in this rather limited scenario, its utility and payback are rather questionable. So, let’s focus on the higher utility cost – and therefore higher potential value – of insights into the operation of HVAC loads at commercial and industrial buildings rather than residences.

Sub-metering could take many forms depending on the data desired. The value in sub-metered data lies in understanding when a load is operating and how much electricity it demands under these circumstances. Several metrics can be monitored when sub-metering electricity usage; for example, kW (true power), kVA (apparent power), kVAR (power factor), voltage and frequency.

Keep in mind that the more metrics that are desired, the more expensive the metering solution. In many cases, simply measuring kW provides a wealth of information just by answering the questions of “when” and “how much” posed above.

A reasonable frequency to report sub-metered HVAC loads, balanced against the overhead of gathering, transmitting and storing the data, is usually in terms of aggregated five-minute periods. Monitoring at much more frequent increments may provide slightly more insights into exactly when such loads transition operation (e.g. exactly when a second stage compressor started once the first stage was already running), but such information can typically be gleaned even from five-minute buckets. In some cases, even aggregating across 15-minute periods can provide sufficient insight into how an HVAC load is operating.

Now that we have identified the desire to examine HVAC load electrical demand and consumption across different time interval periods, what insights can be gleaned?

Understanding HVAC Behavior

Our goal is to both confirm expected behavior as well as identify anomalies for both the group as well as for individual loads. So, let’s examine a sampling of such behaviors and how sub-metered HVAC load data provides facility managers and energy analysts with the data that can allow them to determine if corrective measures are required.

Examining daily peaks: By aggregating HVAC loads and examining when they peak, potential issues can be identified. For example, one would expect that theaters will peak during their busiest times; e.g., Friday evenings and Saturday and Sunday afternoons. If HVAC loads peak during mornings on those days, it may indicate inappropriate scheduling of loads, which is causing them to start cooling far too soon -- or with initial occupancy setpoints that are far too aggressive. Instead, one could slowly ease HVAC loads into operation, shifting peaks to later in the afternoons or early evenings (and then use demand management solutions to help reduce such peaks). Such analysis can also be performed on individual HVAC loads to identify problems if only a few loads are starting too early or running too late in the day.

Examining each HVAC load’s operation over time, particularly in baseline mode of operation (i.e., no demand management controls are in play), can lead to the analysis of properly operating versus faulty HVAC loads. For example, over the duration of days with warm temperatures, noting HVAC loads that only indicate blower activity could indicate either broken compressors or setpoint overrides that are far too warm for normal conditions. Similarly, on cooler days, seeing HVAC loads that peak out during the day when free cooling – or no cooling – is required could indicate that maintenance is needed to correct a faulty economizer damper or an unnecessarily aggressive low set point.

Examining when HVAC loads operate can lead to confirming reasonable scheduling – or indicate scheduling errors. For example, noting loads that operate overnight when the building is unoccupied, especially those that not only operate blowers but also compressors, immediately indicates potential schedule optimizations in terms of tightening up overnight activity. In some cases, such loads may need to operate at reduced levels during unoccupied mode; e.g., venting kitchen areas or allowing cooling for overnight cleaning/stocking crews, since having insight into correct HVAC operation is important for personnel safety and comfort.

Examining the maximum demand of HVAC loads can lead to confirming reasonable operation of these loads.

For example, under hot weather conditions, HVAC loads occasionally require all stages of cooling. If, over the course of weeks of such weather, some loads are observed to only be consuming a portion of their expected maximum demand, say, only 60%, this likely implies that the load’s upper stage or stages are never operating. This could be simply due to the setpoint being so high that upper stages are never requested, and the zone is then not being cooled sufficiently, or perhaps the upper stage compressors are not functioning due to a variety of mechanical faults. Regardless of the reason, simply monitoring peak load anomalies identifies problems that require investigation.

The analysis that sub-metering provides is especially useful if available remotely over the web; i.e., without the need to visit the site and extract the data from a building automation system (if one even exists at the site). Web-based reporting means that reports can be generated and made available as frequently – or as infrequently – as desired, depending on the preferences of the personnel involved.

Encycle provides such HVAC visibility reporting services using the sub-metered data inherent in its Swarm Logic® demand management solution, which is widely used by commercial, industrial and institutional building customers. For more information, please visit encycle.com.

By Mark Kerbel of Encycle Corporation