ASHRAE Journal - November 2020 - 59

COLUMN ENGINEER'S NOTEBOOK

After all, this was based on a sketch. I had not seen the
fan curves, had no idea where the fans were operating
on their curves, no idea where the surge line was and no
idea if we would cross it when we went into the two fan
operating mode.
I will never forget working with the operators to manually set up the scenario; bringing a second fan online
and making educated guesses at the settings we needed
Then, we tripped a fan off and held our breath. There
were no alarms. Granted, we ended up at less than
design ﬂow with one fan running. But that ﬂow was
sufﬁcient to prevent a crash, allowing the cleanroom to
remain in production. Subsequently, we demonstrated
that manually speeding the fan up restored design ﬂow,
all without crashing the fab. Adding a sensor and some
control code allowed the process we had demonstrated
manually to be automated.

N + 1 and Energy Efﬁciency

At the time, I thought this strategy would save energy;
after all, the afﬁnity laws said that running two fans
to serve a given load at reduced speed will save energy
compared to what it will take if you run one fan at full
speed to do the same thing. Unfortunately, my thinking
was abusing the afﬁnity laws.
I was "clued in" when the kW readings from the VFDs
showed higher consumption for the new operating
mode. Contemplating what we had done in the context
of Equation 1 revealed my oversight.


Flow × Static
kW = 
× 0.746
 6, 356 × ηFan × ηBelts × ηMotor × ηVSD

(1)

where
kW = Input to the system to produce the ﬂow and
static pressure
Flow = Flow rate, cfm

Static = Fan static pressure, in. w.c.
6,356 = A units conversion constant that is good for
air at approximately 0 ft to 2,000 ft (mean sea
level) and between -40°F and 120°F (-40°C
and 49°C)
ηFan = Fan static efﬁciency
ηBelts = Belt efﬁciency; well-adjusted V-belts typically
have an efﬁciency of 97% to 98%
ηMotor = Motor efﬁciency
ηVSD = Variable speed drive efﬁciency.
0.746 = Horsepower to kW conversion constant
If you consider the terms in the numerator, the ﬂow
term was unchanged from what it was for single-fan
operation. And because both fans were connected to
a common duct system (other than for some minor
dynamic loss differences associated with splitting
the ﬂow at the fans vs. having all of the ﬂow go to one
fan or the other, and reducing the ﬂow by 50% in the
fan discharge stack) the static term was not signiﬁcantly changed. Thus, energy savings would be solely
due to improvements in the efﬁciency terms in the
denominator.
For these fans (and many fans and pumps I have
looked at where one machine was selected to serve the
full load), the overall efﬁciency did not improve. Equation
2 (which is based on Equation 1) and Figure 3 illustrate this
for both operating modes.
Fan Efficiency. The afﬁnity laws assume that when
you reduce the fan speed, you are moving down a system curve that runs through the selection point. But,
in this case, the operating point shifts horizontally
across the curve to a signiﬁcantly less efﬁcient location (Figure 3a).
Motor Efficiency. Motor efﬁciency improved slightly
due to the slight rise in the motor efﬁciency curve.
However, this could have gone the other way depending
on where you started from (Figure 3b).

Before:



28, 000 × 6.0
kW = 
× 0.746 × 1Fan = 28.25 kW

 6, 356 × 0.802 Fan × 0.9800 Belts × 0.9252 Motor × 0.9600VSD 

After:



14, 000 × 6.0
kW = 
× 0.746 × 2 Fans = 34.07 kW

 6, 356 × 0.669 Fan × 0.9800 Belts × 0.9350 Motor × 0.9440VSD 

Added Cost:
(34.07 kW − 28.25 kW)) = 5.82 kW: For 8,760 hours per year = 50,983 kWh at 3.8 cents per kWh=$1,937
N OVEM BER 2020

ashrae.org

ASHRAE JOURNAL

(2a)

(2b)

59
https://www.ashrae.org/
ASHRAE Journal - November 2020

Table of Contents for the Digital Edition of ASHRAE Journal - November 2020
