Pumping Station Desing - Second Edition by Robert L. Sanks, George Tchobahoglous, Garr M. Jones
Подождите немного. Документ загружается.
• For
booster pumping
or
pumping into
a
distribution
system
without
a
reservoir,
V/S can
provide con-
stant
water pressure during varying rates
of
usage
(this
is the
most common application).
•
Pumps
can be
ramped
up (or
down) slowly
so
that
water
hammer
is
reduced.
In
pipelines
so
long that
5t
0
is
more than
5 min
(about
30 km or 20
mi),
surges
caused
by the
startup
of C/S
pumps
can be
serious. Solid-state starters
or
automatically operated
control valves
can
also reduce water hammer,
but
only
in
shorter pipelines.
The
full-speed operation
of
C/S
pumps
for
long periods (more than several min-
utes)
at
near
shut-off
head causes excessive
(1)
wear
on
pump bearings
and (2)
flexing
of the
shaft.
Disadvantages
of
Variable-Speed
Pumping
The
selection
of V/S
pumping units should only
be
made
after
due
consideration
of the
disadvantages.
•
Except
for
engine drives,
V/S
adds high cost
and
complexity,
requires more equipment
and
more
maintenance,
and
reduces reliability. Troubleshoot-
ing
and
repairs require personnel with on-the-job
training.
Each type
of V/S
drive requires
a
different
kind
of
specialized training.
• If the
operating personnel
do not
understand
the
equipment
or find it
difficult
to
maintain, they
are
likely
to
turn
off the V/S
drive, operate
the
pumps
manually
at
C/S,
and
thus
effectively
lose
the
investment
in
expensive equipment plus
the
advan-
tages
it
offers.
Pump sumps designed
for V/S
pump-
ing
have
insufficient
storage
for C/S
pumping,
and
the
frequent
starting
may
burn
out the
motors
and
certainly reduces their
life.
• V/S
drives have less electrical
efficiency
than
C/S
motor drives (although
V/S
drives
may use
less
energy because
of
lower pipeline
friction
losses).
•
Avoiding vibration
is
more
difficult
with variable
drive
shaft
speeds because
the
"windows" (the
ranges
of
allowable supporting structure
frequen-
cies)
for
avoiding resonance
are
narrower. This
avoidance
should always
be by
design
—
never
by
luck
(see Chapter 22).
•
Instrumentation
for
regulating pump speed
must
usually
be
more
refined,
accurate,
and
costly than
is
required
for C/S
pumping.
An
exception
is the
pneumatically
operated liquid rheostat, which
requires
no
instrumentation.
• V/S
drives
are not
well adapted
to flat
system
H-Q
curves
because
(1)
good
efficiencies
cannot
be
obtained throughout
the
speed range;
(2)
small
changes
in
speed produce large changes
in
dis-
charge;
and (3)
energy losses (and costs)
are
high.
•
Rapid changes
in
technology
can
make particular
models obsolete.
• V/S
drives
may be
much noisier than
C/S
drives.
•
AFDs
in
particular
are
more vulnerable
to
lightning
and
electrical disturbances.
The
disadvantages
of V/S
pumping
are
essentially
advantages
for C/S
pumping.
To
complete
the
com-
parison, consider
the
following:
• The
cost
of the
large
wet
wells required
for C/S
pumping
may
exceed
the
cost
of V/S
drives,
but
note that excavation costs
are
site dependent.
•
Pump sizes
and wet
well design
can be
easily coor-
dinated
so
that
the
number
of
pump starts
per
hour
(for
C/S
pumps)
is
well within acceptable limits.
• A
wide range
in flows can
also
be
accommodated
by
using several
C/S
pumps,
and the
programmers
for
alternating
C/S
pumps
are
more
reliable
and
less
expensive than control systems
for V/S
units.
A
penalty
for
using
C/S
pumps
is a
larger station,
more pumping units, more piping
and fittings, and
discrete increments
of flow.
•
Surges
due to
starting
or
stopping
one
pump
are not
as
severe
as
those caused
by a
power failure, which
may
occur when several pumps
are
running
and for
which
the
station
and
pipeline must
be
protected
from
water hammer anyway. Solid-state starters
or
pump-control valves
can
eliminate surges
in
usual
circumstances.
If the
pumping station discharges
a
small proportion
of the
treatment plant
flow, the
sudden change
of
influent
flowrate due to the
start-
ing
and
stopping
of a
single pump
may be
negligi-
ble.
The
effect
cannot
be
quantified,
so
this
is a
matter
of
judgment.
•
Relatively constant water pressure
from
a
booster
pumping
station
may be
obtained with
C/S
pumps
either with
flat H-Q
curves
(if
suction pressure
is
constant)
or
with elevated
or
pressurized tanks.
Summary
If
objective evaluation
of
these
advantages
and
disad-
vantages leads
to a
decision
to use
V/S, select
the
best
drive
for the
particular application.
15-2.
Design
Considerations
Firm pumping capacity
is the
maximum station dis-
charge with
the
largest pump
out of
service. Similarly,
a
station with
V/S
pumps must
be
able
to
accommo-
date
any flow
(from
minimum
to
maximum)
and
oper-
ate in a
normal manner when
any one of the
pumps
or
drives
is out of
service.
A
V/S
pumping station that includes
C/S
pumps
can
provide
the
same results
as a
station with
all V/S
pumps
—
but
only
if the
relationship between
the
sizes
of the two
kinds
of
pumps
is
correct (see Sec-
tions
15-5
and
15-10)
or if the
sump
is
made large
enough
to
permit
V/S
pumps
at low
speed
to
cycle
on
and
off.
Pumps
The
number
of
pumps
to be
installed
in a
station
depends
on (1) the
range over which
the
influent
flow
rate varies,
(2) the
available pump selections,
and (3)
the
initial
and
operating costs.
Wet
Wells
For V/S
pumping,
the wet
well need only
be a
sump
for
pumping
at the
instantaneous system
flowrate.
Because storage
as
such
in the wet
well
is not
needed
at
all,
the wet
well size
can be
reduced
to the
geome-
try
necessary
for
adequate pump inlet conditions
and
access
—
nothing
more. Furthermore,
the
change
of
water
surface elevation
can be
reduced
to the
amount
needed
for
regulating
the
speed
of the
pumps. Most
V/S
control systems start
and
stop
the
pumps
and
reg-
ulate speed over
a
range
of
about
0.6
m
(2 ft) of
change
in the
water elevations
in the wet
well.
(An
even
smaller range
can be
used,
but a
more sophisti-
cated control system
may be
required.) Hence, some
static
head
can be
saved
as
compared with
a C/S
pumping
system
in
which
the wet
well level typically
fluctuates
over
a
range
of 1.2 m (4 ft) or
more.
It
is
easier
to
design good
wet
wells
for V/S
pump-
ing
than
for C/S
pumping because
the
free
fall
into
the
pool
of a C/S
pumping station
wet
well
can be
avoided
together with
the
entrainment
of air
bubbles
and the
release
of
odors.
The
deposition
of
putrescible solids
is
virtually eliminated
in V/S
pumping, whereas depo-
sition
and
long residence times
of
solids
can be a
problem
in C/S
pumping.
The
design
of
pump sumps
is
discussed
in
Section 12-7.
15-3.
Theory
of
Variable-Speed
Pumping
It may be
helpful
to
review
the
fundamental theory
for
centrifugal
pumps
in
Section 10-3, which
was
explained primarily
by
Equations
10-15
through
10-
21
and by
Examples 10-2, 10-3,
and
10-8.
In
pump manufacturers' published data,
H-Q
curves
are
usually shown
for
impellers
of
several
diameters
at two or
more
speeds.
Such curves drawn
for
pump speeds
of
880, 695,
and 585
rev/min
are
shown
in
Figure
15-1.
At
variable
speeds,
the
pump
actually operates
on an
infinite
number
of
speed
curves between
the
maximum
and
minimum limits.
The
minimum speed
may be
less
than
the
lowest
speed
for
which
a
published curve
is
available.
Most nonclog, dry-well sewage pumps
can be
fur-
nished with
an
impeller
of any
diameter
desired
between
the
minimum
and
maximum shown.
For
example,
if the
pump whose curves
are
shown
in
Fig-
ure
15-1 must deliver
a
maximum
of
0.347
m
3
/s
at
15.2
m
(5500
gal/min
at 50
ft),
the
curve
for
that
impeller would pass through those coordinates
and
would
be
represented
by the
dashed lined labeled
20.1
in.
(impeller diameter).
Speed versus Flow
Curves
When discharging into
a
hydraulic system,
the
pump
must
operate
on the
system head-capacity (H-Q)
curve
at all flows. In
Figure 15-2,
a
typical system
H-Q
curve
has
been superimposed
on the
20.1
-in.
diameter impeller curves.
The
pump delivers
0.133
m
3
/s
(2100 gal/min)
at
10.8
m
(35.5
ft)
when
it
oper-
ates
at 585
rev/min, 0.227
m
3
/s
(3600
gal/min)
at
12.3
m
(40.5
ft) at 695
rev/min,
and
0.347
m
3
/s
(5500
gal/min)
at
15.2
m (50 ft) at 880
rev/min.
A
curve
of
speed versus discharge
can be
plotted
from
these
data,
as in
Figure 15-3. This curve
is
probably
the
most
meaningful depiction
of V/S
pump operation
for
a
single pump. Note that
the
pump
efficiencies
(obtained
from
Figure
15-1)
vary
from
70% at
maxi-
mum
discharge through
83% at
midspeed
to 77% at
low
speed
—
good
efficiencies
throughout
the
usual
discharge rates.
Affinity
Law for
Speed
To
determine
a
speed versus
flow
curve
for any
pump
impeller,
find the flowrates at
which three
or
more
of
the
pump
H-Q
curves (each
at a
different
speed) inter-
sect
the
system
H-Q
curve.
If
only
one or two of the
published pump curves intersects
the
system curve,
points
on
another speed curve
can be
calculated
from
the
affinity
laws
as
shown
in
Example 10-2, Section
10-3. Note, incidentally, that points along
the
system
H-Q
curve
are not
"corresponding
points"
and, hence,
discharge
is not
proportional
to
pump speed.
Figure
15-1.
Pump
characteristic
curves
at
three
speeds.
Minimum
Operating
Speed
The
minimum operating pump
shaft
speed (zero
Q
speed)
is
defined
as the
speed
at
which
the
pump
is no
longer
able
to
maintain discharge against
the
static
head. Hence
it is
unique
for
each pump
and
each
pumping
station,
and it
occurs
at the
speed
for
which
the
pump's
shut-off
head equals
the
static head.
At
zero discharge (zero
Q),
and
only
at
zero
<2,
the
pump
discharge head
can
change without
a
change
in
the flowrate.
Hence,
at
this unique locus
of
points,
Equation
10-16
is the
only
one
needed
for
calculating
the
zero
Q
speed
for a
pump discharging into
a
hydraulic system. Solving that equation
for the
pump
and
system
H-Q
curve
in
Figure
15-2
gives
/TjT
rr^
w
=
/i
L
7
!
= 585
^l
= 501
rev/min (15-1)
^n
2
A/
45
and
a
position
of the
curve
for
this speed
is
shown
as a
dashed
line
in
Figure 15-2. However, most pumps,
especially medium
and
large ones, should
not be
oper-
ated
more than momentarily
at
zero
flow. At
zero
flow
and
at low
discharge rates,
the
pumped
fluid
recircu-
lates
within
the
volute, which causes
an
unbalanced
radial thrust against
the
impeller that results
in a flex-
ing of the
pump
shaft,
which
can
cause bearing
failure
or
fatigue
fracture
of the
shaft
(see Section 10-6).
As
a
rule
of
thumb
it is
usually safe
to
operate
a
pump
continuously
at
discharge rates
of 30% or
more
of
its
best
efficiency
capacity
(bee),
which
is its
dis-
charge capacity
at the
best
efficiency
point (bep)
at the
maximum
recommended speed.
The
pump
manufac-
turer's minimum discharge rate, however, should
always
be
met;
the
minimum recommendation
may be
50% or
more
of
bee
for
some pumps.
By
requiring
custom-engineered pumps (heavier
shafts
and
bear-
ings),
the
problem
with
the
minimum
recommended
operating speed
can be
avoided with most pump
makes.
But a
custom pump, especially
in a
nonclog
design, might
be
expensive,
so
discuss
the
subject
with
one or
more manufacturers. Perhaps
a
larger
number
of
smaller pumps would
be a
good alternative.
Most
V/S
drives provide
a
method
for
preventing
the
pump
from
operating below
a
preselected mini-
mum
speed. When
the
influent
rate
is
less than
the
dis-
Figure
15-2.
Pump
characteristic curves
for a
20.1-in.
impeller.
charge rate
of the
pump
at its
preselected
minimum
speed,
the
pump
is
cycled
on and
off. There
is no
speed variation during
the
on-off
operation.
If the
minimum
influent
rate
is
less than
the
minimum rec-
ommended discharge rate
of a
single pump,
the wet
well
(and
the
sewer pipe
if its
storage capacity
is
uti-
lized) should include
sufficient
storage
to
prevent
cycling
the
pumps
too
frequently
at
this rate. When
the
influent
rate again reaches
the
discharge rate
of the
pump
at its
preselected minimum speed,
the
pump
is
returned
to V/S
operation
by the
control system.
The
following
two
methods
can be
used
to
avoid
cyclic
action
of the
"lead"
pump:
• Use a
larger number
of
smaller
V/S
pumps.
For
example,
use
four
3000-gal/min
pumps instead
of
two
6000-gal/min
units. Select
the
pump capacities
so
that
the
recommended minimum discharge
rate
of
each pump
is
less than
the
minimum
influent
rate.
• Add two
auxiliary
V/S
pumps that operate only
when
the
influent
rate
is
less than
the
recommended
minimum
discharge rate
of the
main
V/S
pumps.
Select
the
capacity
of
each auxiliary pump
to
deliver
a
maximum
flow
that
slightly exceeds
the
recommended minimum discharge rate
of one of
the
main pumps.
Pump
Efficiencies
As
speed changes,
efficiencies
follow
parabolic curves
with
their apexes
at the
origin
of the
graph,
as
shown
in
Figure 15-4. Thus, except
for a
slight
shift
with Rey-
nolds number,
efficiency
is
independent
of
speed. Con-
sequently,
if
pump
H-Q
curves
are
constructed
for
different
speeds,
the
efficiencies
can be
plotted
at the
same time.
By
projecting
efficiencies
from
a
pump
H-Q
curve
along
a
parabola
to the
system
H-Q
curve,
the
efficiency
at any
pump operating point
can be
found.
An
alternative
way to find
pump
efficiencies
along
a
system curve
is to
superimpose characteristic curves
for
selected pump impeller curves
at two or
preferably
three speeds
(as in
Figure 15-2) onto
the
system
H-Q
curve
and
connect points
of
equal
efficiency.
The
pump
efficiencies
along
the
system
H-Q
curve
can
then
be
plotted
as
shown
in
Figure
15-3.
Power
Required
When
the
pump discharges into
a
hydraulic system,
the
power delivered
by the
pump
at any
point
on the
system head curve
is
determined
by the flow and the
head
at
that point (see Equation
10-6).
Figure 15-3. Speed
versus
flowrate
and
efficiency.
According
to the
affinity
laws,
the
ratio
of
power
delivered
by a
pump (water power)
at
"corresponding
points" (such
as
Points
A and B in
Figure 15-4)
at
dif-
ferent
speeds equals
the
cube
of the
ratio
of
speeds.
But
points along
the
system
H-Q
curve
(such
as
Points
A and C) are not
"corresponding
points"
so the
power delivered along
the
system curve does
not
vary
as the
cube
of
the
ratio
of
speeds.
Using Equation 10-6
to
calculate
the
ratio
of
power delivered
at 695 and 585
rev/min
along
the
system head curve
in
Figure
15-4
gives:
(3550)
(40.5)
_
(695V
(2100)(35.5)
1585;
from
which
x =
3.8.
But x
varies along
the
system
head curve.
For the
same pump,
the
exponent
x can
vary
from
less than
2 to
more than
4
along
the
system
curve.
Moreover,
it is the
required pump
shaft
or
brake
power that interests
the
designer.
The
power required
by
the
pump
at any
point
on the
system curve
is
equal
to the
power delivered
at
that point divided
by the
pump
efficiency
at
that point. Pump
efficiency
can
vary
considerably over
the
system curve,
and
this
can
cause
the
pump power requirements
to
further
deviate
from
a
cube relationship. Thus, drive manufacturers'
efficiency
data (which
are
based
on a
cube relation-
ship
of
speed
and
required power)
are not
applicable.
Always
compute power
requirements
from head,
discharge,
and
efficiency
(Equation
10-7)
—
not
from
the
affinity
laws.
15-4.
Pump
Selection
All of the
arrangements described here include
suffi-
cient pumping capacity
to
discharge
the
peak
influent
rate when
any
single pump
is out of
service.
Two-Pump
Facility
For an
installation with
two
pumps, each capable
of
dis-
charging
the
peak
influent
rate,
a
single duty pump with
Figure 15-4. Characteristic curves
for a
single
pump.
Corresponding
points
A,
B,
and E lie on a
parabola
through
the
origin,
as do
corresponding
points
C and D.
Points
A and C (on the
system curve)
are not
corresponding
points.
its
peak discharge rate well
to the right of the bep
pro-
vides
satisfactory
efficiencies
throughout
the
entire
operating range
and
very good
efficiencies
in the
center
of
the
range.
An
example
of the
capability
of a
single
pump
is
shown
in
Figure
15-4.
Note that
the
efficiencies
are
greater than
70%
over
a
range
of flows
equal
to
3:1.
Three
Variable-Speed
Pumps
In
the
usual arrangement
of a
three
V/S
pump system,
two
pumps must
be
capable
of
discharging
the
design
flowrate,
while
the
third
is
idle
as a
standby unit.
Unless
the
system curve
is flat, a
single pump must
be
capable
of
discharging somewhat more than half
of
the
peak
influent
rate.
The
"lead"
pump runs alone
during
periods
of low flow. The
"lag" pump
is
started
when
the
influent
rate slightly exceeds
the
maximum
discharge rate
of the
lead pump,
and
both pumps oper-
ate
until
the
influent
rate again
falls
to
slightly within
the
capability
of the
lead unit.
Assume,
for
example, that
a
pumping station must
discharge
flowrates
from
0.13
m
3
/s
(2000
gal/min)
to
0.50
m
3
/s
(8000
gal/min)
at a
static
lift
of 10 m (33 ft)
and
a TDH at
peak discharge
of 20 m (67
ft).
The
sys-
tem
H-Q
curve
is
shown
in
Figures 15-4
and
15-5.
The
friction head
is
moderate
and the
curve
is
rela-
tively
flat, so a
pump with comparatively steep
H-Q
curves should
be
selected.
When
two
pumps
are
oper-
ating,
the bep
should
be to the
left
of the
operating
point
at
peak
flowrate so
that
the
system
H-Q
curve
sweeps across
the
curves
of
highest
efficiency.
There-
fore,
the bep of a
single pump
at top
speed (880 rev/
min in
Figure 15-4) should
be at the
left
of the
midrange discharge (0.35
m
3
/s
[5500
gal/min]),
and
for
a
single pump operating, only
the
higher
efficiency
curves should straddle
the
system
H-Q
curve.
The
minimum
discharge should
be at
least
30% of the
maximum.
The
pump depicted meets
all of
these
requirements
and
thus appears
to be a
good choice.
Load-Sharing Operation
When
two of the
above pumps operate
in
parallel
at the
same speed,
the
discharge (abscissa)
is
doubled with
no
change
of
head (ordinate),
as is
shown
in
Figure
Figure
15-5.
Characteristics
for two
pumps
(per
Figure
15-4)
in
parallel.
15-5.
(Review Figure
10-24
and the
accompanying text
if
an
explanation
is
needed.) Note that
the
locus
of bep
for
the two
pumps
at
full
speed
is
slightly
to the
left
of
the
peak discharge.
The
efficiencies
throughout
the
full
range
of flows
from
0.13
m
3
/s
(2000
gal/min)
to the
peak discharge, shown
in
Figure 15-6,
can be
kept
between
76 and
83%. Throughout
the
entire range
of
flowrates, the
pumps operate
at 36% or
more
of
maxi-
mum
capacity,
and
neither pump would cycle
on and
off
frequently.
Hence,
the
pump selection
is
indeed
good.
Usually
a lag
pump
is
started when
the
influent
rate
slightly
exceeds
the
capacity
of the
lead pump,
and the
two
pumps then operate
at the
same speed.
The lag
pump
runs continuously until
the
influent
rate
is
again
well
within
the
capability
of the
lead pump.
In the
example
of
Figure 15-6, however,
the lag
pump could
be
stopped
or
started near
the
intersection
of the
effi-
ciency
curves
for
single
and for
double pump opera-
tion.
Staggered
Operation
The
staggered operation method
is
usually employed
to
permit
the use of a
common adjustable-speed con-
troller
for two or
more
V/S
pumps. (Note that
if
there
must
be a
standby pump, there should also
be a
standby
V/S
control system
for the
same
reason
—
to
ensure
reliability.)
The
lead pump operates
on its
max-
imum-speed curve, while
the lag
pump operates
on
reduced-speed curves. Staggered operation
is a
poor
way
to
operate pumping equipment
and
should
not be
used.
The
complexity
of the
controls
is
increased sub-
stantially,
and the
longevity
of the
pumping equip-
ment
is
threatened. Unfortunately, staggered operation
is
sometimes advocated
as a way to
induce
a
designer
to use
variable-speed pumping
at a
"cost
savings."
Read Section 15-5 before deciding
to use a
common
adjustable-speed controller
for two or
more pumps.
The
discharge rates
of the
pumps
at
various
influ-
ent
rates
can be
calculated
as
follows:
• For any
influent
rate
that
exceeds
the
maximum
capacity
of a
single pump,
find the
head,
H',
from
the
system
H-Q
curve.
•
Determine
the
capacity
of the
lead pump
at
H'
from
its
maximum speed curve.
• The
discharge rate
of the lag
pump equals
the
influ-
ent
rate minus
the
lead pump capacity.
A
portion
of
Figure
15-4
is
reproduced
in
Figure
15-7.
As
an
example
of
staggered operation, consider
a
dis-
charge
of
0.38
m
3
/s
(6000
gal/min) shown
as
point
E
on
the
system
H-Q
curve.
The
head
is
16.5
m
(54
ft),
and
at
this head
the
lead pump delivers about 0.33
m
3
/s
(5200 gal/min)
at
maximum speed (point
F). The lag
pump, therefore, must deliver 0.38
-
0.33
=
0.05
m
3
/s
(6000
-
5200
= 800
gal/min)—
also
at
16.5
m (54 ft)
head, which
is
plotted
as
point
G. By
interpolating
Figure
15-6.
Efficiencies
for one and two
variable-speed
duty
pumps
in
load-sharing
operation.
curve
GH
between
the
585-
and
695-rev/min
curves,
the
shut-off
head
is
found
to be
17.4
m (57
ft).
The
shut-off
head
for 695
rev/min
is
19.2
m (63
ft),
so to
produce
a
head
of
17.4
m (57
ft),
the
speed
of the
pump
would
be,
from
Equation
10-16,
/X
1
=
695^17.4/19.2
= 662
rev/min
or
/I
1
=
695^5.1/63
= 661
rev/min
Interpolating between curves
may be
inaccurate
where
the
curves
are far
apart
(as in
Figure
15-7),
but
accuracy
can be
improved
by
using
the
affinity
laws
(Equations
10-15
and
10-16
simultaneously).
The
combined curve
for the
lead pump
(at 880
rev/
min)
and the lag
pump
(at 661
rev/min)
is
found
by
adding
the
abscissas
of the lag and
lead pump curves
(see Figure 10-24
for
further
explanation).
The
efficiency
of the
lead pump operating alone
is
the
same
as
that shown
for the
single pump
in
Figure
15-6.
The
efficiency
of the
lead pump when operat-
ing
with
the lag
pump
is
found
by
projecting
the
head
horizontally
from
point
E on the
system
H-Q
curve
in
Figure
15-7
to an
intersection (point
F)
with
the
maxi-
mum
speed curve (880 rev/min), transferring Point
F
to
Figure 15-4 (note
difference
in
horizontal scales),
and
interpolating between
the
efficiency
lines
to
obtain
an
efficiency
of
approximately
73%.
The lag
pump
efficiency
for a
total discharge
of
0.38
m
3
/s
(6000
gal/min)
is the
efficiency
at
point
G,
which
(transferred
to
Figure
15-4)
is
found
to be
below
the
pump
manufacturer's data.
It is
probably about
55%.
Efficiencies
at
other discharges
are
found
in the
same manner. From these
efficiencies
and the
heads
and
discharges,
a
plot
of
required pump
shaft
power versus
discharge
can be
drawn
as
shown
in
Figure 15-8.
Comparison
of
Operating
Modes
The
power
requirements
for
"load
sharing"
by the
same pumps
as
above
are
also shown
in
Figure 15-8
for
comparison. Staggered operation usually requires
more power than load sharing whenever
the flowrate
varies between
101 and
about 150%
of the
capability
of
a
single pump.
The
influent
rate could remain
within
this range
for
many hours
at a
time, with
a
con-
sequent waste
of
considerable energy.
Of
even more
concern
is the
operation
of the lag
pump
at
very
low
(and
potentially damaging) discharge rates
for
long
periods
of
time.
To
avoid
the
problems
of low
discharge rates,
the
lag
pump
can be
cycled
on and off at a
preselected
minimum
speed whenever
the
influent
rate
is
between
101 and
approximately 130%
of the
maximum output
of
the
lead pump. (The minimum
safe
speed required
for
the lag
pump
to
deliver
its
minimum discharge rate
may
be
quite
different
from
the
minimum speed
required when
the
same pump acts
as the
lead unit).
But
if the lag
pump
is
cycled,
the wet
well and/or
sewer must include
sufficient
storage between
the
proper elevations
to
avoid rapid cycling.
The
only advantage
of
staggered operation
is the
cost saved
by
using
one V/S
controller
for two or
more
pumps.
The net
cost reduction
may be
much less than
the
cost
of one
converter because contactors must
be
added
to
switch
the
converter
to
either motor
and to
Figure
15-7.
Two
variable-speed pumps
in
staggered
operation.
bypass
the
converter
for C/S
operation
of
each motor.
When
the lag
pump
is
started,
the
lead pump
is
locked
into
full
speed
by an
electrical contactor that bypasses
the V/S
controller, which
is
then switched
to the lag
pump.
When stopping
the lag
pump,
the
lead motor
should
also
be
brought
to a
full
stop
—
and
quickly
—
before
the V/S
controller
can be
again connected
to it.
Meanwhile
the
water level
in the wet
well
increases
with
neither pump running.
The
lead
pump must
be
brought
quickly
to an
intermediate
or
full
speed
to
avoid
too
much increase
in wet
well level. Hence,
switching
from
one
pump
to the
other
is
complicated
with
one
controller
and
staggered operation, whereas
with
load sharing there
is a
smooth transition without
auxiliary
switching equipment.
Selections
for
Steep System Curves
If
the
friction head
is
large
in
relation
to the
static
head,
the
best pump selection would
be to
have
the
bep
well
to the
left
of the
single pump maximum dis-
charge rate.
The
pump shown
in
Figure
15-9
would
be
an
excellent selection
to
deliver half
of a
peak
influent
rate
of
0.58
m
3
/s
(9200
gal/min)
with
the
system
H-Q
curve
shown.
Four
Variable-Speed
Pumps
The
same principles
of
selection apply
to
systems
with
four
or
more
V/S
pumps.
A
combination
of V/S
and
C/S
pumps, however,
may
provide
all of the
advantages
of V/S
drives
at a
much lower equipment
cost.
15-5.
Variable-
and
Constant-Speed Pumps
in
Simultaneous Operation
Throughout this
section
it is
assumed that
the
system
H-Q
curve
is flat
(for clarity)
and
that
the
influent
rate
is
always greater than
the
minimum recommended
discharge rate
of one
pump (for brevity).
The
goals
of any V/S
sewage pumping system
are
to
ensure that
• The
station
effluent
rate
is
always equal
to the
influ-
ent
rate
•
Proper operation still occurs with
any
single pump-
ing
unit
out of
service
• V/S
pumps always discharge
at or
above their mini-
mum
recommended rates
•
None
of the
pumps
is
cycled
at
influent rates above
the
minimum discharge rate
of any V/S
pump.
These goals
can be
attained with
a
combination
of
V/S and C/S
pumps operating together
effectively
as
one
large
V/S
unit
if:
• The
"on-line"
V/S
capacity exceeds each step
of
C/S
capacity
to be
added
by 50% or
more
• The
standby pump
has a
complete
V/S
drive.
Many
designers
shudder
at the
cost
of V/S
drives
for
every motor
and try to
"save
money" with
a mix of
Figure
15-8.
Pump
shaft
power
consumption
for two
variable-speed
pumps.