Miles Ryan, P.E., writes a monthly column in Engineered Systems Magazine on Building Commissioning. Read May’s column below:

Introduction

There is a very nuanced and misunderstood situation which plagues a lot of AHUs. It occurs in AHUs which utilize airflow tracking for control of its return fan and either intentionally, or unintentionally, operate in a fully recirculating configuration. This is the first of several articles in a series written with the intent to explain the issue. The series provides several case studies for how I have seen the situation manifested and how a unique solution in each case was needed to rectify the issue.

Scenarios When AHUs Operate in a Fully Recirculating Configuration

AHU sequences of operation often include scenarios where the AHU may be operating in a fully recirculating mode of operation, with the outdoor and relief air dampers closed and the return damper fully open. Common instances of this scenario would include:

• When the AHU first enables, there is often a delay in the opening of the outdoor and relief air damper to allow system to stabilize before introducing outdoor air.
  • This strategy is common in cold climates where freeze-stat trips on initial enabling of the AHU are attempted to be avoided.
• Unoccupied modes of operation where the AHU still operates for temperature/humidity control, but minimum ventilation control is disabled.
  • Such strategies are common in museums and clinics.
  • The fully recirculating configuration would only occur in this mode of operation if the economizer were disabled due to high ambient conditions.
• Unoccupied Heating (also known as Setback) mode of operation; when the AHU enables for a short period of time during unoccupied hours in the winter to blast the building with heat to prevent building temperatures from drifting too low.
• Unoccupied Cooling (also known as Setup) mode of operation; when the AHU enables for a short period of time during unoccupied hours in the summer to blast the building with cooling to prevent building temperatures from drifting too high.
• Warmup mode of operation; when the AHU enables a little before scheduled occupancy for the building and works to raise space temperatures from unoccupied heating setpoints (e.g. 60°F) to occupied heating setpoint (e.g. 68°F) prior to occupants arriving for the day.
• Cooldown mode of operation; when the AHU enables a little before scheduled occupancy for the building and works to lower space temperatures from unoccupied cooling setpoints (e.g. 80°F) to occupied cooling setpoint (e.g. 72°F) prior to occupants arriving for the day.

Such fully recirculating modes of operation can create a very unstable situation if the return fan is controlled via airflow tracking.

Airflow Tracking

Airflow tracking is a return fan control strategy in which the return fan speed is modulated to maintain return airflow at a setpoint. That return airflow setpoint is reset to be an offset below the real-time, measured supply airflow.

• During occupied hours, that offset would typically be the total dedicated exhaust airflow from the spaces the AHU serves, plus some additional amount to ensure slightly positive pressure is maintained for the building in reference to the outdoors.
• During the modes of operation listed above where the AHU is intended to be operating in a fully recirculating manner, it would make sense for the offset used in return fan control to be 0 CFM.

The Unobtainable 0 CFM Offset

Here is the issue. Let’s start off by assuming this is a single-zone, variable air volume AHU (i.e. no VAV boxes). If the outdoor and relief air dampers are fully closed, and the return air damper fully open, the return fan and supply fan are both going to be controlling for the same controlled variable (airflow being recirculated through the unit), but with different sensors. And those sensors (supply airflow measuring station and return airflow measuring station) will never be fully coordinated with each other (i.e. their readings will not consistently match each other exactly). This will result in the fans fighting each other. For the fan whose airflow reading is trending higher, that fan will start to slow down. For the fan whose airflow reading is trending lower, that fan will start to speed up. The death spiral will begin, with one fan ending up at full speed and one fan bottoming out at minimum speed.

What makes this a difficult situation to understand is that multi-zone variable air volume systems are more common, and the supply fan in such a system does not control directly to supply airflow. Rather it is the downstream VAV boxes which dictate the supply airflow and the supply fan controls to static pressure, whose setpoint has been determined to be adequate to ensure the VAVs are capable of having their airflow setpoints satisfied. It is probably best to iteratively demonstrate this recirculating mode death spiral by watching it play out on a BAS graphic.

Figure 1a. Initial operating conditions. Return airflow reading higher than setpoint.

• Figure 1a shows an initial operating scenario where the supply fan is tracking on static pressure setpoint (SA-P SP). The return fan is controlling to active return airflow setpoint (RA-F EFF SP), which is essentially equal to the real-time measured supply airflow (SA-F) as the offset is 0 CFM during this Unoccupied mode of operation. Notice the return airflow measuring station (RA-F) is reading higher than the supply airflow measuring station.

Figure 1b: Return fan slows down in attempt to track on setpoint. Workload shifts to supply fan, who is now no longer meeting supply air static pressure setpoint.

• Figure 1b shows the return fan is starting to slow down to bring return airflow back to setpoint. This shifts more of the workload to the supply fan. The supply air static pressure starts to drift below setpoint.

Figure 1c: Return fan continues to slow down since its airflow is still above its active setpoint and that active setpoint is dropping as well. The supply fan is speeding up more drastically now attempting to regain static pressure control.

• Figure 1c shows the return fan continuing to slow down since its airflow is still above its active setpoint. The supply fan is speeding up more drastically now attempting to regain static pressure control.

Figure 1d: Death Spiral End-state. Return fan is at minimum speed with return airflow still above setpoint. Supply fan is at maximum speed, pulling far more of the work associated with moving air, but still cannot meet static pressure setpoint.

• Figure 1d shows end state. Return fan is at minimum speed with return airflow still above setpoint. Supply fan is at maximum speed, pulling far more of the work associated with moving air, but still cannot meet static pressure setpoint.

In such instances, both fans are operating in inappropriate locations on their fan curves (see Figures 2a and 2b).

Figure 2a: Supply fan operating point. Fan has to overcome frictional losses in return duct when return fan slows down.

Figure 2b: Return fan operating point. Supply fan is creating such a negative suction pressure in the relief air plenum, the return fan operation is akin to pumping water downhill. In some instances, the supply fan can pull more airflow through the return fan than the return fan should ever be capable of moving with that minimum speed fan curve, and the operating point is actually off the chart. What exactly is occurring in such situations is unclear, but I suspect a pressure drop is occurring in lieu of a pressure rise across the fan. VFDs certainly don’t like it, as they often trip out with some very unusual error codes when subjected to this situation.

The fans will stay like this until the Occupied Mode of operation is initiated, or the outdoor air damper opens as part of an economizer sequence. Such scenarios may never be realized by facility operations staff. More often, a fan VFD alarm of some sort is triggered, or a static safety switch (typically an in-unit switch such as supply fan suction) is triggered.

Case Study 1

The first time I encountered this situation was a bit of my own making. I was re-commissioning a clinic AHU. During my initial interview with facility staff, they reported inconsistent building pressure control issues. After reviewing the AHU operation, I speculated the pressure issues were due to their AHUs having the return fan controlling to a speed offset from the supply fan (i.e. return fan commanded speed was always 10% less than supply fan speed). That approach can create inconsistent building pressures since it incorrectly assumes airflow tracks on fan speed. The return airflow does track closely on return fan speed when the return system is of a fixed geometry (i.e. no return valves), but the supply airflow does not track on supply fan speed since the geometry of the supply system is always changing due to modulation of VAV box dampers. This discrepancy can create inconsistent building pressures since supply and return airflows will not track on each other. I recommended they return to the control strategy that was specified when the clinic was built a few years earlier, that of airflow tracking.

The clinic brought back the controls technician who originally programmed the AHU when the clinic was built to assist in implementing this recommendation. The technician warned me they were forced to abandon the airflow tracking strategy soon after the building was built, but he could not remember exactly the reason why.

We pressed ahead and restored the airflow tracking control strategy for the return fan. The AHU had an unoccupied mode of operation that would run the unit as normal, but with the minimum ventilation requirements being reset to 0 CFM. The program utilized a 0 CFM offset for the return fan control during such unoccupied modes of operation. The first night, the outdoor air temperature was warm and the economizer was disabled. The sequence of events described in Figure 1 occurred and the supply fan low static pressure safety switch triggered. This event jump started the controls technician’s memory. He recalled that it was the exact reason they had abandoned airflow tracking a few years prior.

Putting our heads together, we opted to use a fixed speed offset for the return fan during Unoccupied Mode when the economizer was disabled (i.e. a full recirculating mode configuration) but retained airflow tracking return fan control during all other scenarios. This allows for more consistent building pressure control during Occupied Mode of operation and avoided the Recirculating Mode Death Spiral from ever occurring during Unoccupied Mode.

Conclusion

If you have not seen a sequence like the one described in Case Study 1, I would bet you have seen an AHU operating in the situations described in the forthcoming case studies I will detail in the months ahead. This situation with AHUs operating in full recirculation mode actually occurs in far more scenarios than just the modes of operation listed at the beginning of this article. Such situations are often created unintentionally. Stay tuned as I walk through at least 5 more case studies of where the AHU recirculating mode death spiral has reared its head.