QSE Logo


Commissioning: Trend Analysis Summary

This is the final commissioning column (for a while) dedicated to trend log analysis. Since November, 2011, we have looked at a variety of trend logs and gleaned meaningful systems performance information from them. This month I will describe the “issue” disclosed through analysis of the trend graph included in my May, 2012 Commissioning column followed by a wrap-up summary regarding trend logs and new construction commissioning.

May 2012 Trend Log Analysis

Last month’s trend graph showed about two weeks of 2 minute interval data from the building automation system (BAS) for selected points on an office building rooftop unit. The unit had hot water heating and chilled water cooling coils and full air-side economizer capability. It was a variable air volume (VAV) system supplying 55°F air to VAV/reheat terminal units. During the trend period the outdoor air conditions were cool and dry enough for the system to be in economizer mode the entire time.

One of the first challenges in reviewing the trend graph was noting that the hot water trend points, i.e., 0-100%, represented the heating valve’s “percentage closed”. This is intuitively difficult for most of us to grasp and adds an unnecessary level of complexity to the analysis process. For the most part, we have been trained to think of “0%” as fully closed and “100%” as fully open. I believe any direct digital control system worth having will allow the programmer to present valve and damper positions consistently as (0-100%) = (Closed-Open). As commissioning professionals, I recommend that we review the control system specification to confirm that this is a requirement for programmers, regardless of which brand BAS is selected.

With only a few exceptions, the chilled water valve is signaled to be fully closed (0%) throughout the entire trend period. The hot water valve is signaled to modulate between 100% (fully closed) and about 80% (20% open) in its attempt to maintain the discharge air temperature at its setpoint of 55°F. With no source of cooling between the mixed air and the discharge air, why is the discharge air temperature lower than the mixed air temperature? We can only surmise that the “closed” cooling valve is leaking past and lowering the air temperature such that some heat is actually required to keep the air temperature at 55°F.

This is a condition which would likely go un-noticed without trend log analysis, especially in an existing building. There are no temperature control problems at the rooftop unit, as it is consistently maintaining the desired 55°F discharge air. However, both chilled water energy (electricity) and heating hot water energy (natural gas) is being wasted.

Finally, this commercial office building rooftop unit is clearly operating continuously day and night, weekdays and weekends. When asked about this, the building operators said this unit could not shut down because it served a data closet that had a 24/7 cooling load. This prompted evaluation of an energy conservation measure to cool the data closet with a stand-alone split system, thus allowing the rooftop unit to be shut down during unoccupied hours.

Trend Analysis Summary

During new construction commissioning, trend log analysis is not a substitute for field functional performance testing. However, I highly recommend analyzing trend logs and addressing performance issues identified through trend logging as a pre-test commissioning activity. It will improve the chances of the field functional performance testing being successful and everyone’s testing time being used as efficiently as possible.

In addition, pre-test trend analysis would only be the first step in functional performance testing HVAC systems. Field testing would still be needed to focus on things that wouldn’t have been caught in the trend logs, such as:

  • Other-season modes of operation – i.e., if left to run on its own, the HVAC system will only respond to the weather/occupancy conditions at the time of trending. In most climates above the Maxon-Dixon line, that means there are probably two other weather conditions that would need to be simulated in the field.

  • Safety devices (freezestats, high static switches, etc.)

  • Failure mode operation, i.e., what happens if a pump fails?

  • Extreme load conditions, i.e., seeing components (boilers, chillers, valves, etc.) sequence on/off based on all possible load conditions

  • Occupant-driven load conditions – i.e., demand ventilation, occupancy sensors, etc.

  • Integration with other systems

  • Performance of systems/equipment not monitored/controlled by the central automation system


Commissioning: Getting It Right

Engineered Systems, June 2012

Rebecca Ellis, PE, LEED AP BD+C, CCP, CPMP, CxA
Questions & Solutions Engineering
1079 Falls Curve
Chaska, MN  55318