ASHRAE Journal - May 2021 - 47

COLUMN ENGINEER'S NOTEBOOK

Control Strategy-Option 2
Having recognized transportation lag, a solution
becomes obvious: control the fans based on their
cold basin leaving water temperature. This cuts the

100

Valve Command (%)

FIGURE 4 Cascading instability.

Temperature (°F)

to the chiller, are full of 65°F (18°C) water.
Now, imagine the load drops. Since the entering temperature to the condenser is a steady 65°F (18°C), the
drop in load will show up as a reduction in leaving water
temperature. We'll assume the leaving water temperature drops and stabilizes at 68°F (20°C). This would happen gradually in the real world. To make things easier to
visualize, we'll assume it shows up as a step change, i.e.,
if you were watching the temperature with a magical,
massless thermometer, you would see the water leaving
the chiller drop 1°F (0.6°C) in a fraction of a second.
When the load reduction occurs, the pipe between the
chiller and the tower is full of 69°F (21°C) water, and it will
take about 1.6 minutes for the ﬁrst parcel of cooler, 68°F
(20°C) water to reach the tower and fall through the ﬁll
to the cold basin. Since the sensor controlling the process
does not know the load changed (because at the instant it
happened, the cold basin and piping to the sensor location were full of 65°F [18°C] water), it has continued to
operate the fans at a speed that generates a 4°F (2°C) temperature drop. As a result, the parcel of water that entered
the hot basins at 68°F (20°C) is cooled to 64°F (18°C).
It will be another 2.4 minutes before the parcel of
cooled water reaches the temperature sensor controlling
the process. System reaction will be further delayed by
the thermal mass impact discussed previously.
By the time the control system realizes the load has
dropped, several minutes will have passed, and the process will have over cooled and undershot the targeted
setpoint. Recovery will take a while because all that time
the cooling tower was ﬁlling the piping leaving the cold
basin with colder than desired water.
The transportation lag will make the departure below
setpoint appear to persist, causing the proportional plus
integral (PI) control process to compensate by shutting
down the fans and opening the bypass valve. For the system under discussion, that happened frequently.
Because the bypass valve was oversized, the reaction
produced was out of proportion to the need, triggering further instability, which rippled out to the system's
chilled water side, where load changes produced would
feed back into the condenser water system (Figure 4).

0
12a

Chiller Water Return Temperature
Chiller CW Supply Temperature
Secondary Loop Chilled Water
Control Valve Command

Chilled Water Supply Temperature
Chiller CW Return Temperature
CW Bypass Valve Command

transportation lag in half for a signiﬁcation portion of
the system (Table 1).
But, like most engineering solutions, solving one problem leads to others.
* Two control loops = two setpoints. The setpoints need
to be coordinated so if one is changed, so is the other.
* With two temperature sensors involved, the " out of
the box " accuracy comes into play.
Even the best sensors have a manufacturing tolerance; two sensors with identical accuracy speciﬁcations
may not read the same value when subjected to the
same condition. For applications like our case study, the
relative accuracy of one sensor compared to the other
becomes critical. Ensuring relative calibration is performed along with training the operating team in that
regard will be an important part of the commissioning
process for this option.

Making it Happen
Conceptually, the logic required by the two options is
illustrated in Figure 5. For those interested in the details,
pneumatic and DDC logic diagrams for both options are
at https://tinyurl.com/ASHRAECWDetails, along with a
more detailed version of the system diagram.
M AY 2021

ashrae.org

ASHRAE JOURNAL

https://www.tinyurl.com/ASHRAECWDetails http://www.ashrae.org

ASHRAE Journal - May 2021

Table of Contents for the Digital Edition of ASHRAE Journal - May 2021