Beruflich Dokumente
Kultur Dokumente
The pumping volume is lubricated with vacuum oil, which means that in
reality:
To improve p one should increase the pumping volume and decrease
the dead volume. The best way is to reduces the back pressure well
below 760 Tr e.g., by multistage pumping.
s
c
V
p =760 +p (oil v.p.)
V
Example
A simple piston pump has a displacement of 1L (V
p
). It is connected to a
chamber of 10L (V
0
). If pumping start at 1atm, what will be a pressure in
the chamber after 4 complete cycles?
In a single cycle of the pump operation the number of particles (n
0
) in the
pumped volume will be reduced to (n
1
)
P N ( )= P 0 ( ) 1
V
D
V
|
\
|
.
|
N
For four complete cyclesP(4) = 0.656 atm.
( )
( )
( ) ( )
0
0 0 0
0 1 0 1 0
1
0
0 0
1 0 0 0
0 0 0
; ; ;
1
p
p
p
p p
PV
PV P
n n n n n V V
RT RT RT
RT
P n
V
V
P P RT
P V V
PV
V V P
RT
nR
V
T
V V
= = A = =
=
=
A
| |
= = =
|
\ .
Diaphragm pumps
A flexible metal or polymeric diaphragm seals a small volume
at one end.
At the other end are two spring-loaded valves:
one opening when the volumes pressure falls below the
outside pressure,
other opening when the volumes pressure exceeds the
outside pressure.
A cam ( ( on a motor shaft rapidly flexes the diaphragm,
causing gas transfer in one valve and out the other.
Diaphragm pumps often have two stages in seriesto
produce a lower vacuum, or in parallel, to produce a higher
pumping speed.
In general, diaphragm pumps have low pumping speeds (<10
clm) and produce a poor ultimate vacuum (1 Torr to 10 Torr).
They do exhaust into the atmosphere and their low costs
make them attractive roughing pumps.
Diaphragm pumps are also used for simple vacuum filtration,
thin film evaporation, distillation, gel drying applications, and as
sample movers for gas analyzers and sample extraction.
Dry Vacuum Pumps
Roots Blowers / Booster Pumps
Precision shaped rotors assistant to housing and to each other to within 0.05-0.25mm
Rotors spin at 2500 to 3500 rpm.
Gears synchronize the rotors. Must be below 10 Torr to operate.
Roots pump is excellent for moving large quantities of
gas at higher pressure.
The lobes in cross section are similar to the figure
eight.
They counter-rotate to continuously transfer gas from
one side of the pump (inlet) to the other (outlet).
Air enters from the inlet and the volume is
discharged in the (C) and (d) positions to the outlet
The compression ratio of the pump is small and
varies with the molecular weight of the gas.
-light gases easily escape back into the vacuum
vessel around the edges of the rotors, creating a
beck stream,
- rotary or piston pumps must back roots pumps.
Typically pump speed: N~1500-3000 rpm,
Gap (clearance) between the rotors: 1-0.15 mm
Roots Blowers / Booster Pumps
Isaiah Davies invented the design principle of the rotary
lobe blower (Booster Pump) in 1848, but it was some
twenty years later before Francis and Philader Roots
applied it in practice in the US, first as a water wheel in
their woolen mill.
IMPORTANT! There is no oil sealing in the gas-handling
part of this pump
Roots Blowers / Booster Pumps (cont)
The theoretical pumping speed of a Roots pump is
S
th
=4*n*V,
where n=rotational speed of the rotors, V=volume of gas isolated from inlet
The theoretical gas throughout (Q
th
) is
Q
th
=S
th
*P
in
BUT! Internal leakage can occur because of the
clearances within a Root pump, reducing the amount of
gas handled
Effective gas throughout (Q
eff
)is
Q
eff
=S
th
*P
in
- Q
int leak
Q
int leak
=C
gap
(P
out
-P
in
)+S
ad
*P
out
,
C
gap
= conductance of the internal clearances
S
ad
=volume rate of flow of gas transported by
adsorbtion-desorbtion by the rotors
(P
out
-P
in
)= pressure difference across the Root pump
1) inlet, 2) Exhaust, 3) water jacket 4) screw, 5) oil, 6) gas path, 7) timing
gears, 8) bearings 9) shaft seals, 10) oil seal
This pump works by pushing gas from the inlet (1) along the gas path (6) to
the exhaust (2).
Because the screws are interleaved, the gas is trapped in pockets formed by
the outer-wall and the screws.
Dry Screw Pump
Schematic of scroll pump
L. Creux, US patent no. 801,182.
The pump consists of arcs
of circles of the
same pitch.
Cryo-condensation
Sorption Pumps
Gases are pumped by
Cryocondensation: gases freeze onto cold solid
surfaces of zeolite (See L1 saturated vapor)
Cryosorption: gases are trapped in a porous zeolite
Zeolite: originates from greek Zein (to cool)+ Lithos
(stone)
Zeolite ->Dehydrated aluminosilicate: e.g., Na2O Al2O3
nSiO2mH2O
Dehydration creates channels with calibrated diameter
size, 0.1 nm<D<10 nm, and the network of chambers, of
the volume of the order of 1000 A
3
, interconnected by
narrow channels.
The effective surface ~ 1000 m2/g! 50% of the zeolite
volume is constituted by chambers and channels.
Cryopump
Cooling gases to the extent that gas molecules lose sufficient
energy to form condensation layers.
A cryogenic surface will trap any molecule that contacts the
surface if it is cold enough.
Equilibrium Vapor Pressure
Equilibrium vapor pressure is the state where as many
molecules are condensing as are vaporizing. Equilibrium
occurs when the rate of gas molecules returning to the
liquid/solid (condensing) is equal to the rate of energetic
molecules becoming gaseous (vaporizing).
What determines the Pressure inside a Cryopump?
Surface Equilibrium
When the number of molecules arriving on the
chamber surface (adsorbing) equals the number
leaving the surface (desorbing), then the system is in
Surface Equilibrium
Equilibrium Vapor Pressure:
- CONDENSATION
- VAPORIZATION
Surface Equilibrium:
- ADSORPTION
- DESORPTION
Equilibrium
Vapor Pressure
Cryocondensation
At T=4.2 K Helium is still boils
Cryosorption
Cooling gas molecules to the extent that gas molecules,
upon contacting a sufficiently cooled surface, lose
enough energy to accumulate on the surface.
A flat cryoadsorbing plate retains some molecules and
allows molecules to continue moving.
Cryosorption
Sieve material, such as charcoal, provides greater surface
area and limited apertures.
Increased surface area provides greater capacity.
Released molecules remain confined.
Irregular surface constricts motion.
Cryosorption of hydrogen, neon, and helium accomplished.
The sorption pumps are equivalent in function to the rotary oil pump.
They are used for rouging vacuum system to a pressure level of 10
-1
-10
-5
Torr.
Cryosorption pumps are used most often as starting devices for pumping large volumes
of gas from atmospheric pressure.
Cryosorption pumps have a simple blow-off valve this makes their regeneration
automatic.
Cryosorption pumps do not back-stream they are low cost and, almost problem-free.
Vessel is cooled by immersion in liquid nitrogen (LN
2
) 77 K.
Each sorption pump requires about 10liter of LN
2
and about 20 minutes to cool down.
Several sorption pumps are often combined on a manifold.
Pumps must be regenerated by heating to 250 C for 30 mins. to melt frost and degas
the molecular sieve material.
Sorption Pumps
Aluminum body
Porous zeolite
Neck flange
LN
2
Bucket
LN
2
Sorption Pump Bank Design
Banks of 2 or more sorption pumps are often staged together
Faster pumping speeds are achieved by connecting the pumps in parallel
Lower base pressures are reached by operating a bank of pumps in series
Staging of sorption pumps
Single length sorption pump
Sorption Pump
Ranges
Single Stage
10
-2
torr
Double Stage
10
-4
~ 10
-5
torr
Accessories
Nitrogen Dewar
Bakeout Sleeve
Sieve Material
Image of used sorption pump with dewar
and heater sleeve
Image of used sorption pump
with sieve material port opened
Cryosorption and Cryocondensation
Air gases and water vapor still condensed -noncondensible gases captured.
Cryopump Concept
Cryopumps are designed to create these
condensing and adsorbing surfaces.
The first stage of the cold head (9)
cools the thermal radiation shield (5)
and the baffle (6) of the pump.
The second stage of the cold head
(7) is used to cool the cryopanels (8).
Operating temperature of 10-20K are
attained and led to cryocondensation
of N2 O2 and Ar
Cryopump : Configuration
The active pumping surfaces are made of Cu-the high thermal conductivity material ,
and they are tightly linked thermally to second stage of the cold head
H2, He, and Ne are also adsorbed on to these surfaces which are partly covered
with activated carbon
Cryogenic Pumps
This gas capture pump is particularly useful for systems containing gases other
than helium or hydrogen.
Their major use at present is in sputtering systems applied to semiconductors
processing, where oil-free operation and a huge capacity for pumping argon
process gas are needed.
They are particularly suited to pumping high molecular weight gases in the 10
-6
to 10
-9
Torr range.
The most important feature of the cryopump is the cleanliness of the vacuum.
The pumping speed can be vary high and the ultimate vacuum excellent.
In contrast with ion pumps, which are also oil free the cryopumps, is mach less
susceptible to failure or damage if switched on at high pressure.
There are two major classes of pumps
Liquid Pool
-Liquid helium temperature
(~4K)
Closed cycle
-Refrigerator (~12K)
Supplemented by cryosorption
Cryopumps
Use a closed-loop helium cryogenic refrigerator.
Gases are pumped by two processes:
Cryocondensation (H
2
O, CO
2
, N
2,
O
2
, Ar,
solvent vapors)
Cryosorption (H2, He, Ne)
Primary parts are:
Compressor
Expander
Cold Head
Typically pumping speed 100 - 1000 L/s.
The pump has two low temperature zones, an inner surface held at approximately 20K
and a surrounding surface at approximately 80K (Helium refrigeration).
The inner surface is coated with activated carbon that assists in pumping hydrogen by
adsorption. The low temp are achieved by attaching the pump directly to a helium cryo-
compressor.
Cryopumps
Potential Problems:
Must be regenerated to extract the trapped gases
Allow to warm to room temperature (slow), or
Use a built-in heater to warm to 250 C and outgas (fast).
Regeneration takes the pump off-line for several hours.
Regeneration process can produce considerable pressure.
Pumps have a safety pressure relief valve on the bucket.
Must be started from below 100 mTorr
Use a mechanical roughing pump
Rotary Vane, Oil-Sealed Mechanical Pump
Typical housing for a wet
rough pump that is of the
rotary mechanical pump type
inlet exhaust
discharge
outlet valve
rotor
vanes
single stage
two stage
high vacuum
stage
rough vacuum
stage
inlet
exhaust
discharge
outlet valve
Rotary Vane Pump
Clockwise Expanding volume of gas
Half revolution.
Expanding gas seals off the inlet port
Gas starts to be
compressed
Gas ballast is admitted.
Completion of one cycle
Gas and ballast air are
compressed, gaining enough
pressure to force open the
discharge valve.
The end of the discharge
portion of the cycle
Since both vanes operate, in one rotation of the rotor a volume of gas equal to twice that
indicated in fig. b is displaced by the pump.
The volume rate at which gas is swept round the pump, S, is
S = 2V*n
where V is the volume between vanes A and B (fig. b), and n is the number of rotations per
unit time (usually 350-700 r.p.m.).
In theory, the lowest pressure achieved by the pump is determined only by the fact that the
gas is compressed into a small but finite "dead volume". When the system pressure
becomes so low that, at maximum compression, the gas pressure is still less than that of
the atmosphere it cannot be discharged from the pump
The compression ratio of the order of 105 are required to produce pressures of the order of 10
-
2
Torr.
The lowest {ultimate) pressure achieved by a single stage rotary pump is about 5 x 10
-3
Torr.
Parallel connection of two identical rotor-stator systems will provide twice the displacement
but the same ultimate pressure. A two-stage pump may reach 10
-4
Torr
The oil in an oil sealed pump serves three important functions:
providing a vacuum seal at the pump exhaust,
as a lubricant
provides cooling for the pump.
Oil rotary pumps can cause significant contamination of HV and UHV systems!
Rotary vane pumps
Rotary vane pump typically consists of a steel cylinder mounted on a driving shaft
Its axis is parallel to the axis of the stator, but is displaced from this axis, such that it makes contact with the top surface of the stator,
the line of contact lying between the two ports. A clearance of 2-3 microns must be at these points
The whole of the stator-rotor assembly is submerged in a suitable oil.
As vane A passes the inlet port (fig. a), the vacuum system is connected to the space limited by the stator, the top seal, the rotor and
vane A. The volume of this space increases as the vane sweeps round, thus producing a pressure decrease in the system. This
continues until vane B passes the inlet port (fig. b), when the volume of the gas evacuated is isolated between the two vanes.
Further rotation sweeps the isolated gas around the stator until vane A passes the top seal (fig. c). The gas is now held between vane
B and the top seal, and by further rotation it is compressed until the pressure is sufficient (about 850 Torr) to open the exhaust
valve, and the gas is evacuated from the pump.
Since both vanes operate, in one rotation of the rotor a volume of gas equal to twice that indicated in fig. b is displaced by the pump.
Thus, the volume rate at which gas is swept round the pump, referred to as pump displacement S, is
S = 2Vu
where V is the volume between vanes A and B (fig. 2b), and n is the number of rotations per unit time (usually 350-700 r.p.m.).
The contacts of the vanes and rotor with the stator form three separate chambers each containing gas at different pressure.
These contacts must therefore make vacuum-tight seals, especially for the top seals which must support more than one
atmosphere pressure difference. For this reason the inner surfaces- of the stator, that of the rotor and vane, are very carefully
machined.
In theory, the lowest pressure achieved by the pump is determined only by the fact that the gas is compressed into a small but
finite "dead volume". When the system pressure becomes so low that, at maximum compression, the gas pressure is still less
than that of the atmosphere it cannot be discharged from the pump. Subsequent pumping action reexpands and recompresses
the same gas without further decreasing the pressure in the system.
The ratio of the exhaust pressure to the inlet pressure is termed the pump compression ratio. Thus, to produce pressures of the
order of 10
-2
Torr, pumps having compression ratios of the order of 105 are required. Oil also performs the function of filling the
dead volume, thus increasing the compression ratio. The lowest {ultimate) pressure achieved by a single stage rotary pump is
about 5 x 10
-3
Torr. Parallel connection of two identical rotor-stator systems will provide twice the displacement but the same
ultimate pressure. A two-stage pump may reach 10
-4
Torr
Rotary vane pumps
Gas ballast to improve
pump operation
When condensible vapors are pumped,
they can only be compressed to their
saturation vapor pressures
Example: H
2
O at 70C will condense
when compressed to p
s
=312 mbar;
further compression will create no
increase in pressure,
so an overpressure of the outlet valve
will not be achieved
Water would condense into the oil;
solvents could do worse!
A gas ballast (air or N
2
metered into
the pump) will reduce the partial
pressure of the condensing gas so that
it can be pumped out of the system
gas
ballast
Gas ballast
The gas ballast increases the pressure in the pump chamber just before
compression starts.
The compression ratio reduced and vapors which might otherwise have
condensed may remain in the vapor phase.
The required amount of gas ballast is
S
0
= inlet pumping speed [Ls
-1
]
Q
ballas
=0.1*S
0
*p
0
= amount of ballast entering the pump
P
o
atmospheric pressure
P
vap,1
=pressure of vapor (1) in the gas mixture to be pumped
P
1
= the pressure of non-condensable gas at the inlet
P
s
= saturated vapor pressure of the vapor being pumped at the temperature of
the coolest part of the pump (about 60-70C)
P
ex
= the pressure at the pump exhaust
.1
1
1
1
vap
ballas o ex s
s vap
p
p
Q S p p
p p
(
| |
= + ( |
|
(
\ .
Gas ballast
Vapor tolerance is the maximum pressure of vapor that rotary- vacuum pump
can take in and transport at T=20C, p=1013mbar
Vapor tolerance for water vapor WVT is
P
w,a
=partial pressure of water vapor in the ballast air
,
0
0.1 *
s w a
ex s
p p
WVT p
p p
Example
Vacuum chamber was cleaned with dematerialized water and drained. A rotary-vain
pump (S
0
=25m
3
h
-1
) was connected to chamber to remove the residual water from
the chamber.
Calculate the maximum amount of water that may be handled by the pump
If the exhaust T=75C and air used as the gas ballast is of T=20C, humidity RH50% and pressure
p
0
=1013mbar.
It is known that saturated vapor pressure for water at 75C is p
s
=385.5mbar and at 20C
p
s
=23.4mbar
Pay attention ! RH=50%, so p
w,a
=23.4/2 mbar
,
0
385.5 23.4/ 2
0.1 * 0.1*1013 37.3
1400 385.5
s w a
ex s
p p
WVT p mbar mbar
p p
| |
= = =
|
\ .
The maximum amount of water that can be handled is
Q=S
0
*WVT=25/3.6 Ls
-1
*37.3mbar =259mbar*Ls-1
Q=pV=mRT/M
1 1
1 1
1 1
259 18
0.16 0.58
83.14 348
mbarLs ggmol
m gs kgh
mbarLmol K K
= = =
Accessories used with mechanical pumps
Foreline sorption traps -- help control oil backstreaming (transport of pump oil
through the inlet line into the system being pumped
usually filled with molecular seive materials
may be cooled or heated
Anti-suckback valves -- prevent oil backstreaming in the event of a power failure
A fast-acting normally closed valve is placed at the pump inlet; if power is lost,
the valve automatically closes
vacuum systems hate
power failures!!
HIGH VACUUM PUMPS
Oil Diffusion Pump (DP)
The name diffusion pump does not describe the operation of the pump accurately.
Diffusion pump is a vapor jet pump which transports gas by momentum transfer on
collisions with the vapor steam.
Diffusion pumps were first conceived and constructed
by W. Gaede (1915-Germany) and I. Langmuir (1916-
U.S.A).
They operate on the principle of transferring
momentum from high velocity vapor molecules to the
gas molecules that are to be moved out of the system.
The vapor molecules are formed by heating a suitable
condensable fluid.
The early pumps used mercury for this purpose.
In the late 1920s, it was found that high molecular
weight oils having high boiling points and low vapor
pressures could be used as
pumping fluids.
Today all diffusion pumps use some form of oil.
Pump Construction
Nozzles
Hot dense vapor is forced through central jets angled downward to give
a conical curtain of vapor.
Gas molecules are knocked downwards and eventually reach the fore
vacuum pump (Rotary)
The oil pool at the bottom of the
pump is heated, causing oil
vapor to be forced up the jet
stack.
The vapor strikes the
umbrellas, and is projected
downward and outward through
the nozzles of the jet stack.
In passing through the narrow
jets, the oil vapor flows at a
velocity near that of sound.
The jet streams are in series; the illustrated pump is
a "three-stage" type.
Pumping Speed
10
-10
10
--3
10
--1
P
u
m
p
i
n
g
S
p
e
e
d
(
A
i
r
)
1 2 3 4
Inlet Pressure (Torr)
Critical Point
1. Compression Ratio Limit
2. Constant Speed
3. Constant Q (Overload)
4. Mechanical Pump Effect
In normal operating range DP is a constant speed device
Oil Diffusion Pump
Since the pump works by momentum
transfer, the compression ratio depends
on the atomic mass. The thermal velocity
of light gas is much greater, so the
molecules are pumped less efficiently.
Gas Compression Ratio
Pumping fluids
Apiezons : A - natural hydrocarbons p(25 C) ~ 10-5
B - destileted hydrocarbons p(25 C) ~ 10-7
C - destileted hydrocarbons p(25 C) ~ 10-8
Silicon:
Dc703 - phenylmethyl dimethyl cycioxane [(CH3)3SiO] [(CH3)3SiO] [(CH3)3SiO]
Dc704 - tetramethyl tetraphenyl trisiloxane 4 pieces chain
Dc705 - pentaphenyl trimethyl trisiloxane 5 pieces chain
Santovac (Convalex) - synthetic polyphenol ether space
lubricant expensive!
Liquid N
2
reservoir with baffles
How the LN2 Trap Works
LN
2
in a container in a room doesnt go
poof into gas
holds itself at 77 K and all evaporate due to
finite heat of vaporization of LN
2
is 5.57
kJ/mole, one liter of cryogen would last for
LN
2
4.5 hours if i the thermal load on the
inner shield is 10 W
Gas Vapor Pressure (mbar)
(H
2
O)
(Ar)
(CO
2
)
(CO)
(He)
(H
2
)
(O
2
)
(Ne)
(N
2
)
10
-22
500
10
-7
>760
>760
>760
350
>760
760
Water cooled baffle
Oil is contained and condensed by the diffusion
pump. A small amount can escape toward the
HiVac area of the system. This "backstreamed" oil
is detrimental to the system.
To contain it, a water-cooled baffle is located
between the diffusion pump and the cold trap.
Most back-streamed diffusion pump oil
molecules are condensed on the internal water
cooled baffle disc and returned to the diffusion
pump in the form of liquid oil.
Oil Diffusion Pump
Potential Problems:
Backstreaming of oil vapor can occur if forepressure becomes too
large.
Backstreaming occurs for pressures of 1 to 10 mTorr.
Cold cap on top of multistage jet assembly helps to reduce this.
Liquid nitrogen filled cryotrap also helps to reduce this.
Maximum tolerable foreline pressure (critical forepressure) must not
be exceeded, or pump will dump or blow-out, sending oil up into
the chamber.
Pump can overheat if cooling water fails
Most pumps have a thermal cutout switch.
Pumping requires low vapor pressure oil
Water, dirt, or other impurities will raise vapor pressure.
Only special oils are suitable for diffusion pump use.
Example
Diffusion pump has a pumping speed of 1000Ls
-1
at p
in
,10
-3
mbar, falling to
400Ls
-1
at p
in
=10
-2
mbar.
The critical backing pressure p
crit
for the pump is 0.4mbar!
a) Calculate the minimum size (pumping speed) of the baking rotary pump if diffusion
pump is in at 10
-2
mbar
At 10
-2
mbar the throughput of the diffusion pump is
(Q
max
)
in
= p
in
*S
effDP
=10
-2
mbar*400Ls
-1
=4mbarLs
-1
(Q
max
)
in
=(Q
max
)
back
=S
eff
*p
crit
Approximate
value of the pumping speed of baking pump is
S
eff,back
=Q
max
/p
crit
=4mbarLs
-1
/0.4mbar=10Ls
-1
b) Calculate the minimum size (pumping speed) of the baking rotary pump if diffusion
pump is in at 10
-3
mbar.
(Q
max
)
in
= p
in
*S
effDP
=10
-2
mbar*1000Ls
-1
=1mbarLs
-1
S
eff,back
=Q
max
/p
crit
=1mbarLs
-1
/0.4mbar=2.5 Ls
-1
Turbomolecular Pump
The turbomolecular pump was invented in 1958 by Becker
Turbomolecular pump imparts a directed momentum to gas
molecules through a rapidly rotating blades.
Turbine blades rotating at high speed transfer momentum to gas molecules to force
them out the exhaust (must be backed).
9,000 to 90,000 rpm motor speeds!
- 20 to 60 blades per disk
The pump is characterized by a compression ratio and ultimate pressure.
Work only in molecular regime, use after roughing pump is spent (< 100 mTorr)
Typically 100 to 800 L/sec pumping speeds
UHV can be readily achieved (better if used in combination with a titanium
sublimation pump).
Pumping speed is proportional to the rotor speed and is constant when normal
rotation speed is achieved.
Turbo pumps have advantages over diffusion pumps:
no back-stream oil into the vacuum system at any time
can be started and stopped in a few minuets
can be directly connected to the chamber without a high vacuum valve.
!!!But turbo pump can be noisy and they induce vibration, and the compression ratio
for hydrogen and helium are low
Turbomolecular Pump
The compression ratio of the turbine establishes the base pressure.
The compression ratio is higher for higher molecular weights:
Approximately: log10K = 1.5 (M)
1/2
, K-compression ratio
For H
2
, M = 1, so K = 10x1.5 = 30 = very small
For hydrocarbons, M = 100, so K = 1015 = very large
Base pressure is usually limited by H
2
.
Turbomolecular Pump: Structure
10 to 20 stator/rotator pairs assembled
on a single shaft.
The blades rotate at 20,000-90,000 rpm.
The stators are between the rotors in
order to direct the gas molecules.
Rotor stator distance is less than 1 mm.
The pairs near the inlet are for high
throughput whereas those near the exit
are at low pumping speed.
Mechanically rotor bearings and
magnetically suspended bearings.
No trapping gas.
No uncertainty about whether the pump is
pumping.
If the rotor is turning, it is pumping.
Traps, baffles are not required between the
vacuum chamber and the pump inlet.
Molecule
V
Moving Wall with Speed V
Schematic of the rotor and stator of a turbomolecular pump
Inlet
Outlet
R
o
t
a
t
i
o
n
b
l
a
d
e
s
N
o
n
-
R
o
t
a
t
i
o
n
b
l
a
d
e
s
Gas captured by the upper stages is pushed into
the lower stages and successively compressed
to the level of the fore-vacuum pressure.
As the gas molecules enter through the inlet, the
rotor, which has a number of angled blades, hits
the molecules. Thus the mechanical energy of
the blades is transferred to the gas molecules.
With this newly acquired momentum, the gas
molecules enter into the gas transfer holes in
the stator. This leads them to the next stage
where they again collide with the rotor surface,
and this process is continued, finally leading
them outwards through the exhaust.
The relative velocity between the rotating blades
and stator blades makes it probable that a gas
molecule will be transported from the pump inlet
to the pump outlet
Each blade is able to support a pressure
difference. Because this pressure ratio is small
for a single stage many stages are cascaded
Rotation axis is in the center of the
rotor and in the vertical direction.
The rotor moves at a high speed
such that when a molecule hits a
blade , momentum is imparted to
the molecule in a direction almost
perpendicular to the blade surface.
Turbomolecular Pumps
Potential Problems
Very high speed rotor blades have close-mating stator blades.
Slight imbalances can cause vibration and bearing wear problems.
Sudden blast of atmospheric pressure can bend the blades down, causing
catastrophic failure, crashing the pump.
Lubrication of the high speed rotor is an engineering problem.
Circulating oil is most reliable, but pump must be right-side-up.
Grease-lubricated bearings are less reliable, but allow pump to be placed at
any orientation.
Too high of a pressure will cause aerodynamic lift and drag.
A mechanical foreline pump must be used
Aerodynamic lift can bend blades, causing catastrophic failure.
A high voltage combined with a magnetic field causes electrons to travel in a
helical path with an energy sufficient to ionize gas atoms.
The ions are accelerated so they strike a Ti cathode and become buried in the
plate.
Sputtered Ti coats everything: tubes, plates, and pump casing.
Several pumping mechanisms are possible, including chemical reaction (getter
action(, ion burial, and neutral burial (the last two accounting for the pumps
ability to pump inert gases).
Can be started below 10
-6
mbar!!!!.
Pumping speed is gas dependent and drops off below 10
-9
mbar.
A little less expensive than turbo pumps.
Ion pump for UHV (and high HV)
Single cell from ion pump
Cathode Anode
Ion pump for UHV (and high HV)
Diode Pump: High pumping speed for getterable gases low speed for noble gases (no getter action).
Argon instabilities when buried Ar is released from cathodes at high gas loads.
Noble (or Differential) Diode Pump: The two cathodes are made of different materials, Ti and Ta
Ta is also a getter (slightly less active). When ions impinge on Ta, the higher mass of Ta results in
more energetic neutrals, which penetrate deeper (less Coulomb interaction), both in anode and pump wall.
Triode Pump: Glancing incidence collisions with Ti grid cathode increase Ti sputter yield and yield
of neutrals, both leading to increased burial.
Ion pump for UHV (and high HV)
A anode B- magnetic flix line produced by a
permanent magnet attached to the outside of the
pump,K1,K2- cathode plates made of Ti , in
contact with pump walls
1- electron ( ) in a spiral track,
2- gas molecules ( ), ionized gas particles( )
3-sputtered Ti ( ), 4- implanted buried, gas
particles, 5- gas particles incorporated into
sputtered Ti
Ion pump for UHV (and high HV)
Diode-type sputtering ion pump-pumping mechanism
A sputtering ion pump consist of a large number of such cells
Ions generated in cell impinge on the cathode plates and sputter Ti. The amount
sputtered is roughly proportional to the pressure in the pump. This material
adheres to the surface of anode, opposite cathode and getters reactive gases.
Some ionized particles penetrate well into the cathode.
This mechanism of pumping is effective for all types of ion
As cathode sputtering proceeds such particles are released giving rise to gas
pressure
Ion pump for UHV (and high HV)
Triode-type sputtering ion pump-pumping mechanism
A anode, B- magnetic flux line produced by a
permanent magnet attached to the outside of pump,
K- Ti cathode in the form of grid, behind K is a
collector target plate F at anode potential
1- electron ( ) in a spiral track,
2- gas molecules ( ), ionized gas particles( )
3-sputtered Ti ( ), 4- implanted buried, gas particles,
Triode-type sputtering ion pump has a grater
efficiency for pumping noble gases
It is due to geometry of the system favors grazing incidence of the ions from the
discharge on the cathode grid. This increases the sputtering yield and the number of
ions that are neutralized and reflected. These ions travel to the target plate with high
energy and implant themselves but! With negligible sputtering.
Ions which are not neutralised cannot move against the electrical field between
K and F and return towards K
The enhanced pumping speed for noble gases is because the sputtering is less at F
and the reemergance of implanted particles is reduced
Ion Pumps
Pump life depends on quantity of gas pumped> 20 years at 10
-9
mbar
The nominal pumping speed of ion pump is given by the max
in the pumping speed curve for air
Gas H
2
H
2
O CH
4
O
2
Ar He
Relative
pum..speed
%
145 80 100 80 25 30
Example
A chamber is fitted with a triode ion pump with pumping speed S
air
=100Ls
-1
A gas mixture containing
5%CH
4
+10%O
2
in N2 is admitted. A required total pressure in the chamber is 10
-6
mbar
-Calculate the flow rate of the gas mixture .
Ptot=P
CH4
+P
O2
+P
N2
=1*10
-6
mbar= Q
CH4
/S
CH4
+Q
O2
/S
O2
+Q
N2
/S
N2
=
=5*10
-2
*Q
tot
/S
CH4
+0.1*Q
tot
/S
O2
+0.85*Q
tot
/S
N2
S
N2
=S
air
=100Ls
-1
, S
O2
=0.8S
air
, S
CH4
=S
air
, (see the table above)
1*10
-6
mbar=5*10
-4
Q
tot
+1.25*10
-3
Q
tot
+8.5*10
-3
Q
tot
=10.25*10
-3
Ls
-1
Q
tot
mbar
Q
tot
=9.76*10
-5
mbarLs
-1
Getters and Gettering
Gettering is the process of removing impurities from the
active track regions
Physical getters are in cryostats where a zeolite material is used to physically absorb and hold water vapor.
If a zeolite material is placed at liquid nitrogen temperature it will absorb and hold gases.
Chemical getters provides a pumping action by a chemical reaction where a chemically active gas combines with a
chemically active metal to form a low vapor pressure solid compound.
The chemically active metal can be an element or an alloy that can be called a getter metal (GM)
These reactions with active gases are fairly straightforward.
They all form a low vapor pressure ceramic compound
The active gas is permanently removed from the vacuum chamber.
The inert gases are not pumped at all due to they are inert and will not react with GM
Hydrogen (H
2
) does not react to form a chemical compound but merely dissolves in
GM to form a solid solution.
How To Use Getters
There are two separate categories of chemical getters:
i- evaporable, ii-non-evaporable.
Evaporable getters are used in electron tubes:
when the tube is pumped down a slug of getter metal is
heated to a high temperature to evaporate it so that it
can subsequently condense on the tubes inner surfaces
to forms a high surface area reactive coating that will
remain as an in situ pump after the tube is pinched off.
Titanium sublimation pump (TSP) : heating
titanium/molybdenum alloy filament to 1450
o
C Ti
sublimes and then condenses into a thin film on the
inner surface of the chamber
The lifetime of Ti film is dependent upon the gas load it is required to pump
The pumping speed of is directly proportional to the total surface area of the film.
The pumping speed will decline as it reacts the gas in the chamber
The Ti source will become saturated with H
2
when it cooled after sublimation
temperature, and re-heating releases the hydrogen.
How To Use Getters
Non-evaporable getters (NEG) remain in the solid state.
Usually it is zirconium alloys in any solid form: chunks or pellets or thin films
bonded to metallic substrates.
Getter Activation process is required for NEGs:
- NEG exposed to air has skin of oxides, nitrides, etc over its surface.
-NEG bulk will be saturated with dissolved H2.
NEG will be essentially inert and will not provide getter-pumping surface.
Getter Activation the process to prepare the getter surface for pumping.
It is done in situ by heating under vacuum after being installed.
During heating skin layer will diffuse into the NEGs bulk and the H
2
will be
driven out of solid solution
The time and temperature required for activation will vary with the specific
getter alloy,
The vacuum level required is at least 10
-4
torr
H
2
remaining in the chamber will be re-pumped by the getter,
Gases will be removed from the vessel by reaction on the NEGs surface until enough gas
is reacted to once again skin over the surface and reacted skin begins to act as a
barrier for further dissolution of H
2
.
The higher temperature will result in pumping and diffusion into the bulk taking place at
the same time as sort of a continuous activation cycle.
Depending upon the actual NEG alloy in use, a compromise temperature must be chosen
that trades-off diffusion rate for H2 retention.
Re-activation is required each time the NEG material sees air or any other active gas .
When an active gas encounters a hot NEG material, an exothermic reaction occurs that
can run away in terms of temperature and present a very real safety problem.
In a cycled system, the chamber can be released to argon instead of an active gas, or the
getter can be contained within a separate housing that can be isolated from the
chamber by an appropriate valve.
How To Use Getters
Getter Pumps
An important class of getter pumps are the Non Evaporable Getters (NEGs)
These are alloys of elements like Ti, Zr, V, Fe, Al which after heating in vacuum
present an active surface where active gases may be gettered
Traditionally, the getters take the form of a sintered powder either pressed into the
surface of a metal ribbon or formed into a pellet
Getter pumps
In recent times, thin films of getter material have been formed on the inside of
vacuum vessels by magnetron sputtering
These have the advantage of
pumping gas from the vacuum chamber by gettering
and of stopping gases from diffusing out of the walls of the vessels
Titanium Sublimation Pumps (TSP)
a type of getter pump
When a gas molecule impinges on a clean metal film, the sticking probability can
be quite high.
For an active gas with the film at room temperature, values can be between 0.1
and 0.8.
For noble gases and hydrocarbons sticking coefficients are very low (essentially
zero)
Evaporated films, most commonly of titanium or barium, are efficient getters and
act as vacuum pumps for active gases.
Example
UHV chamber evacuated by turbomolecular pump (TMP) backed with oil-rotary pump and Titanium
Sublimation Pump (TSP). A total pressure of 4*10
-9
mbar is achieved and this consists of 50%Ar+50%H
2.
. On reaching this pressure the valve to the TSP is closed and pressure in chamber rise up to 8*10
-
9
mbar.
- Calculate the new partial pressure of H
2
in the system, taking in account that Noble gases will not be
pumped by TSP
P
tot
=P
Ar
+P
H2
=2*10
-9
mbar(P
Ar
)+2*10
-9
mbar(P
H2
)
AP=8*10
-9
mbar -4*10
-9
mbar=4*10
-9
mbar
new P
H2
=2*10
-9
mbar+4*10
-9
mbar=6*10
-9
mbar
- Calculate the pumping speed of TSP for H
2
, if the effective pumping speed of TMP is 100Ls
-1
Initially P
tot
=4*10
-9
mbar=P
Ar
+P
H2
or
For Ar P
Ar
=Q
pv,Ar
/100Ls
-1
Q
With TSP closed off
With TSP open
2
, , 9
, , ,
4 10
pv H pv Ar
eff TMP eff TMP eff TSP
Q Q
mbar
S S S
= +
+
2 2
2
7 1
, , , 9
1 1 1
7 1
,
2 10
;8 10
100 100 100
6 10
pV Ar pV H pV H
tot
pV H
Q Q Q
mbarLs
P mbar
Ls Ls Ls
Q mbarLs
+
= = +
=
7 1 7 1
9
1 1
,
7 1
9 1
,
1
,
2 10 6 10
4 10
100 100
6 10
2 10 ; 200
100
eff TSP
eff TSP
eff TSP
mbarLs mbarLs
mbar
Ls Ls S
mbarLs
mbar S Ls
Ls S
= +
+
= =
+
Starting the system ( all off):
1.start BP
2.open V1
3.open DP water
4.wait until G1 reads below 10
-1
Tr
5.turn on DP heater
6.fill lq. Nitrogen trap
7.wait until the DP warms up (20-40 min.)
DP diffusion pump
BP primary (backing) pump
V1 backing valve
V2 bypass valve
V3 - vent (air inlet) valve
V4 high vacuum valve
G1 fore-vacuum gauge
G2 high vacuum gauge
OPERATING A DIFFUSION PUMP VACUUM SYSTEM
Pumping down (initial pressure in the chamber
760 Tr, all in the pump start-up positions):
close V3
close V1
open V2
wait until G2 reading goes below ~ 10-2 Tr
close V2
open V1
slowly open V4
wait until G2 reads the desired pressure
DP diffusion pump
BP primary (backing) pump
V1 backing valve
V2 bypass valve
V3 - vent (air inlet) valve
V4 high vacuum valve
G1 fore-vacuum gauge
G2 high vacuum gauge
Venting the chamber:
close V4,
open V3
Stopping the system.
pump down the chamber.close V4
turn off DP heater
stop lq. Nitrogen refill
wait 20-40 min for the DP to cool down
close V1
stop BP
close DP water
OPERATING A DIFFUSION PUMP VACUUM SYSTEM
189
UNEP 2006
Training Agenda: Pumps
Introduction
Type of pumps
Assessment of pumps
Energy efficiency opportunities
190
Assessment of pumps
Pump shaft power (Ps) is actual horsepower
delivered to the pump shaft
Pump output/Hydraulic/Water horsepower (Hp) is
the liquid horsepower delivered by the pump
How to Calculate Pump Performance
Hydraulic power (Hp):
Hp = Q (m3/s) x Total head, hd - hs (m( x (kg/m3( x g (m/s2( / 1000
Pump shaft power (Ps):
Ps = Hydraulic power Hp / pump efficiency Pump
Pump Efficiency (Pump):
Pump = Hydraulic Power / Pump Shaft Power
UNEP 2006
hd - discharge head hs suction head,
- density of the fluid g acceleration due to gravity
191
UNEP 2006
Assessment of pumps
Absence of pump specification data
to assess pump performance
Difficulties in flow measurement and
flows are often estimated
Improper calibration of pressure
gauges & measuring instruments
Calibration not always carried out
Correction factors used
Difficulties in Pump Assessment
PUMPING SYSTEM CHARACTERISTICS
STATIC HEAD
Resistance of the system which is to be overcome to initiate the flow is called head.
"Head" may be simply defined as any resistance to the flow of a pump.
Static head is the difference in height between the source and destination of the pumped liquid
Static Head can be divided into two parts :
Static Suction Head (Hs) : Also called the suction lift .It is the vertical distance from surface
of liquid at source (can be positive or negative)
Static Discharge Head (Hd) : Vertical distance between pump centreline to destination
STATIC HEAD
STATIC HEAD
VERSUS FLOW
PUMPING SYSTEM CHARACTERISTICS
FRICTION HEAD
The pressure drop due to resistance of flow in the pipe and fittings is called the
friction head
Friction head is effected by the flow and increases with increase in flow
FRICTIONAL HEAD
VS FLOW
SYSTEM HEAD
VS FLOW
TOTAL DYNAMIC HEAD (TDH) OR SYSTEM HEAD
TDH = STATIC DISCHARGE HEAD STATIC SUCTION HEAD+FRICTION HEAD
PUMPING SYSTEM CHARACTERISTICS
PUMP PERFORMANCE CURVES
Head and flow determine the performance of a pump
PUMP OPERATING POINT
The pump operating point is determined by the intersection of the system curve
and the pump curve
PUMP PERFORMANCE CURVE PUMP OPERATING POINT
Pump Curves
For a given pump
1. The pressure produced at a given flow
rate increases with increasing impeller
diameter.
2. Low flow rates at high head, high flow
rates at high head.
3. Head is sensitive to flow rate at high flow
rates.
4. Head insensitive to flow rate at lower
flow rates.
Pump Curve
- used to determine
which pump to
purchase.
- provided by the
manufacturer.
Pump Curve
Pressure
increases
with
diameter
Low flow at
high head
Head
sensitiv
e to flow
at high
flow
rates
CHARACTERISTICS CURVES
T
O
T
A
L
H
E
A
D
(
M
)
FLOW RATE(LPM)
%
A
G
E
E
F
F
I
C
I
E
N
C
Y
S
H
A
F
T
P
O
W
E
R
CENTRIFUGAL PUMP PERFORMANCE CURVES
Total head reduces with increase in flow rate
Power required increases with increase in flow rate
Efficiency increases with increase in flow rate and reaches peak before falling
down
DSM measures to reduce
consumption in pumps
Demand side management measures
Use of friction less foot valve.
Use of HDPE pipes
Use of appropriate capacitor.
Use of higher size suction pipe compared to
delivery pipe.
Overall Efficiency = Hydraulic power ( P2) X
100/ Power input ( P1)
Pump efficiency. = Hydraulic power ( P2) X
100/ Power input to pump shaft ( P3)
Hydraulic Power ( P2) = Q X Total Head ( hd -
hs ) X p X g / 1000
Q = discharge in m/s
Pump Performance calculation
p = density of fluid in kg/ m
g = acceleration due to gravity ( m/s)
P1 = 1.732 X V X I X pf
P3 = P1 X eff.of motor.
Key Parameter for determining
efficiency
Flow
Head
Power
Flow Measurement Techniques
Tracer Method
Ultrasonic flow measurement
Tank filling method
Installation of online flow meter
Determination of total head
Suction head
measured from pump inlet pressure gauge reading
Discharge head
This is taken from the pump discharge side Pr.
gauge
Pumps: (Monoblock Pumps used for Drinking
water)
Sl.No ITEM Pump1 Pump2
1 Make Beacon Beacon
2 Capa hp/kW 5 (3.7) 5 (3.7)
3 Pipe sizes S/D 65/50 mm 65/50 mm
4 Head 25 m 25 m
Typical name plate details of pump
5 Discharge 8 L/s 8 L/s
6 Amps 8 8
7 Overall Efficiency 47% 47 %
Typical name plate details of
pump
PUMP EFFICIENCY CALCULATION
SHEET
SL.NO. ITEM Pump1 Pump2 Units
1 Voltage 406 407 volts
2 Current 6.0 6.6 amps
3 P.F 0.83 0.82
4 Power input 3.5 3.48 KW
Typical performance calculations
5 Dis. Head 22.2 22.2 m
6 Suc.Head 2.2 2.2 m
7 discharge 8 8 L/s
8 Hyd.Power 1.57 1.57 KW
9 Pump Effi. 44.85 45.10
Typical performance
calculations
1. Replace the existing discharge line with
the 50 mm dia pipes to reduce the friction
losses.
2. Provide water level switch to switch off
the pump sets as soon as the Tank is filled.
3. Replace the existing pump sets with high
efficiency pump sets.
Typical observations/
Recommendations
Size pump correctly
Operate close to the best efficiency point.
Size all piping and valves correctly
Avoid all leakages.
Factors to be considered from user
side
Flow control Strategies
Varying speed
Pumps in parallel
stop/start control
Flow control valve
By pass control valve
Trimming impeller
Use of VFDS
Energy conservation opportunities
in pumping
Operate pump near best efficiency point.
Replace old pumps by energy efficient pumps
Reduce system resistance by pressure drop
assessment and pipe size optimization.
Provide booster pump for few areas of higher
head.
Energy conservation
opportunities in pumping
Conduct water balance to minimize water
consumption.
Ensure availability of instruments like
pressure gauges, flow meters.
Repair seals and packing to minimize water
loss.
Avoid valves in discharge side as far as
possible.
Operate pumpset during non-peak hours.
CHARACTERISTICS CURVES
SYSTEM CURVE : System curve is the change in flow
with respect to head of the system.
PUMP PERFORMANCE CURVE : is the change in flow
with the total head developed by the pump
BEST EFFICIENCY POINT (BEP)
When system curve is superimposed on pump curve it
crosses at the pump BEP
BEP is the pumping capacity at maximum impeller
diameter ,or the point at which the efficiency of the
pump is highest
Obtaining exact flow rate as per BEP is not practically
possible
Flow rate deviation from BEP results in loss of
performance and pump instability
Flow rate can be matched corresponding to the BEP by
following methods :
a) Adjusting the pump speed
b)Adjusting the pump impeller diameter
c) Changing the impeller design
d)Adjusting the system resistance
e)Modifying static head
f) Providing a system bypass flow route
Pump Specification
Recall Mechanical Energy Balance
( )
2
4
2
2 2
V
K
D
L
f
p
z g
V
W
i
(
+ +
A
+ A +
A
=
o
kg
m N -
( )
c
i
c c
g
V
K
D
L
f
p
g
z g
g
V
W
2
4
2
2 2
(
+ +
A
+
A
+
A
=
o
m
f
lb
lb f t -
Both equations describe work that must be supplied to system
Pump Head
What happens if the MEB is multiplied through by g (g
c
/g)?
( )
(
+ +
A
+ A +
A
=
2
4
2
1
2 2
V
K
D
L
f
p
z g
V
g g
W
i
o
What are the units (SI)?
2
2
m
s
kg
m N -
2
2
2
3
m
s
s kg
m kg
-
-
= m =
W/g has units of length and is known as the pump head
^
Example
Tank A
2
Tank B
1
3
Why do we choose point 2 rather than 3 for MEB?
What kind of valve to uses to control flow rate?
Example
Tank A
2
Tank B
1
3
g
V
K
D
L
f
g
p
z H
i
2
4 1
2
(
+ + +
A
+ A =
\
|
+ =
g
V
H
2
2
min
|
Head vs. Flow Rate
|
|
.
|
\
|
+ =
g
V
H H
2
2
min
|
g
p g
z H
c
A
+ A =
min
g
V
K
D
L
f
i
2
4 1
2
(
+ +
Quadratic
In V or q
System Response
What happens when flow control valve is closed?
Resistance (f) increases
Flow rate decreases
Need more head to recover flow rate
Tank A
2
Tank B
1
3
System Response
Valve Open
Constant
Head Response
Constant
Flow Response
Valve Closed
Pump Curves
Pump manufacturers supply performance curves
for each of their pumps. These are normally
referred to as pump curves. These curve are
generally developed using water as the
reference fluid.
The following can be read directly from a pump
curve:
Head vs. flow rate information for any fluid
Pump efficiency for any fluid
Pump horsepower for system operating with water
Pump Performance Curves
Developed
Head
Impeller
Diameter
Efficiency
Flow Rate
NPSH
Horsepower
http://capsicum.me.utexas.edu/ChE354/resources.html
Power Input
For fluids other than water:
q
W
m P
=
|
.
|
\
|
-
-
-
-
|
|
.
|
\
|
-
|
|
.
|
\
|
-
|
.
|
\
|
-
|
|
.
|
\
|
-
=
min
s
hp s
lb f t
f t
lb
gal
f t
min
gal
q
lb
lb f t
g
g
H
hp P
f
m
m
f
c
60 550
48 . 7
1
) (
3
3
q
m
Power Input
Easier Way
fluid
water
fluid
water
fluid
Gr Sp
P
P
. . = =
|
.
|
\
|
+
=
1
2
1
2
4
m
V
opt
=
3
] [
] [
] [
ft lb
s lb m
s ft V
m
m
opt
=
=
=
Remember
Maximize pump efficiency
Power input (hp) should be minimized if possible
Selected impeller diameter should not be largest or smallest
for given pump. If your needs change switching impellers is
an economical solution
NPSH required by the pump must be less than NPSHA
Outline
Pipe routing
Optimum pipe diameter
Pressure drop through piping
Piping costs
Pump types and characteristics
Pump curves
NPSH and cavitation
Regulation of flow
Pump installation design
Piping and Pumping
Learning Objectives
At the end of this section, you should be able to
Draw a three dimensional pipe routing with
layout and plan views.
Calculate the optimum pipe diameter for an
application.
Calculate the pressure drop through a length of
pipe with associated valves.
Estimate the cost of a piping run including
installation, insulation, and hangars.
List the types of pumps, their characteristics, and
select an appropriate type for a specified application.
Draw the typical flow control loop for a centrifugal
pump on a P&ID.
Describe the features of a pump curve.
Use a pump curve to select an appropriate pump and
impellor size for an application.
Predict the outcome from a pump impellor change.
Define cavitation and the pressure profile within a
centrifugal pump.
Calculate the required NPSH for a given pump
installation.
Identify the appropriate steps to design a pump
installation.
References
Appendix III.3 (pg 642-46) in Seider et al.,
Process Design Principals (our text for this
class).
Chapter 12 in Turton et al., Analysis, Synthesis,
and Design of Chemical Processes.
Chapter 13 in Peters and Timmerhaus, Plant
Design and Economics for Chemical Engineers.
Chapter 8 in McCabe, Smith and Harriott, Unit
Operations of Chemical Engineering.
Pipe Routing
The following figures show a layout
(looking from the top) and plan
(looking from the side) view of
vessels.
We want to rout pipe from
the feed tank to the reactor.
Plan View
50 ft
feed
tank
reacto
r
40 ft
steam
head
er
35 ft
piping
chase
60 ft
reactor
feed
tank
piping
chase
50 ft
35 ft
steam
heade
r
30 ft
45 ft
40 ft
10 ft
Layout View: Looking
Down
reactor
50 ft
feed
tank
reacto
r
40 ft
steam
header
35 ft
piping
chase
60 ft
Plan View
= out
= in
reactor
feed
tank
steam
heade
r
30 ft
85 ft
20 ft
60 ft
35 ft
10 ft
10 ft
Layout
View
Pipe Routing Exercise
Form groups of two.
Draw a three dimensional routing for
pipe from the steam header to the
feed tank on both the plan view and
the layout view.
Size the Pump
globe
valve
check
valve
200 ft
150
ft
1. Determine optimum pipe size.
2. Determine pressure drop through pipe
run.
100 gpm
Optimum Pipe Diameter
The optimum pipe diameter gives
the least total cost for annual
pumping power and fixed costs. As
D , fixed costs , but pumping
power costs .
Optimum Pipe Diameter
C
o
s
t
/
(
y
e
a
r
f
t
)
Pipe Diameter
Total Cost
Annualized
Capital Cost
Pumping
Power Cost
Optimum
Example
Two methods to determine the
optimum diameter:
Velocity guidelines and Nomograph.
Example: What is the optimum
pipe diameter for 100 gpm water.
Using Velocity Guidelines
Velocity = 3-10 ft/s = flow rate/area
Given a flow rate (100 gpm), solve for
area.
Area = (t/4)D
2
, solve for optimum D.
Optimum pipe diameter = 2.6-3.6 in.
Select standard size, nominal 3 in. pipe.
Nomograph
-Convert gpm to
cfm 13.4 cfm.
-Find cfm on left
axis.
-Find density (62
lb/ft
3
) on right axis.
-Draw a line
between points.
-Read optimum
diameter from
middle axis.
3.3 in
optimum
diameter
Practice Problem
Find the optimum pipe diameter for 100
ft
3
of air at 40 psig/min.
A = (s/50ft)(min/60 s)(100 ft
3
/min) = 0.033
ft
2
0.033 ft
2
= 3.14d
2
/4
d = 2.47 in
Piping Guidelines
Slope to drains.
Add cleanouts (Ts at elbows)
frequently.
Add flanges around valves for
maintenance.
Use screwed fitting only for 1.5 in
or less piping.
Schedule 40 most common.
Calculating the Pressure
Drop through a Pipe Run
Use the article Estimating pipeline
head loss from Chemical Processing
(pg 9-12).
AP = (t/144)(AZ+[v
2
2
-v
1
2
]/2g+h
L
)
Typically neglect velocity differences
for subsonic velocities.
h
L
= head loss due to 1) friction in
pipe, and 2) valves and fittings.
h
L
(friction) = c
1
fLq
2
/d
5
c
1
= conversion constant from Table 1
= 0.0311.
f = friction factor from Table 6 =
0.018.
L = length of pipe = 200 ft + 150 ft =
350 ft.
q = flow rate = 100 gpm.
d = actual pipe diameter of 3
nominal from Table 8 = 3.068 in .
h
L
due to friction = 7.2 ft of liquid
head
Loss Due to Fittings
K= 0.5 entrance
K = 1.0 exit
K=f(L/d)=(0.018)(20) flow through tee
K=3[(0.018)(14)] elbows
K=0.018(340) globe
K=0.018(600) check valve
Sum K = 19.5
h
L
due to fittings = c
3
K
sum
q
2
/d
4
= 5.7 ft
of liquid head loss due to fittings.
h
Lsum
=7.2 + 5.7 ft of liquid head loss
Using Bernoulli Equation
AP = (t/144)(AZ+[v
2
2
-v
1
2
]/2g+h
Lsum
)
AP = (t /144)(150+0+12.9)= 70.1 psi due
mostly to elevation. Use AP to size
pump.
elevation velocity friction and
fittings
Find the Pressure Drop
check
valve
400 ft
50 ft
400 gpm
water
4 in pipe
Estimating Pipe Costs
Use charts from Peters and Timmerhaus.
Pipe
Fittings (T, elbow, etc.)
Valves
Insulation
Hangars
Installation
$
/
l
i
n
e
a
r
f
t
Note:
not
2003 $
EFFECT OF SPEED
Q N , Flow rate is proportional to rotational speed
H N
2
, Head is proportional to square of rotational speed
P N
3
, Power is proportional to cube of rotational speed
Variable Speed Pumps
Advantage: Lower operating cost
Disadvantage: Higher capital cost
System head requirement
(no valve)
Pump curve
for D
i
H (ft)
q (gpm)
q* (desired)
q produced by pump
with no flow control
RPM
1
RPM
2
Affinity Laws
In some instances complete sets of pump curves
are not available. In this instance the pump
affinity laws allow the performance of a new
pump to be determined from that of a similar
model. This can be useful when modifying the
operating parameters of an existing pump.
Affinity Laws
|
|
.
|
\
|
=
1
2
1 2
D
D
q q
|
|
.
|
\
|
=
1
2
1 2
RPM
RPM
q q
2
1
2
1 2
|
|
.
|
\
|
=
D
D
H H
2
1
2
1 2
|
|
.
|
\
|
=
RPM
RPM
H H
3
1
2
1 2
|
|
.
|
\
|
=
D
D
hp hp
3
1
2
1 2
|
|
.
|
\
|
=
RPM
RPM
hp hp
5 1
2
1
1
2
1
1
|
|
.
|
\
|
=
D
D
q
q
PARALLEL & SERIES OPERATION
Regulating Flow from
Centrifugal Pumps
Usually speed controlled motors are
not provided on centrifugal pumps,
the flow rate is changed by adjusting
the downstream pressure drop (see
pump curve).
Typical installation includes a flow
meter, flow control valve (pneumatic),
and a control loop.
Typical Installation
FT
FC
FV
operator
set-point
Designing Pump Installations
use existing pump vendor, note spare
parts the plant already stocks.
select desired operating flow rate,
maximum flow rate.
calculate pressure drop through
discharge piping, fittings, instrumentation
(note if flow control is desired need to
use pressure drop with control valve 50%
open).
add safety factor to calculated head 10
psig spec pump for 20 psig, 150 psig
spec pump for 200 psig.
using head and flow rate, select impeller
that gives efficient operation in region
of operating flow rate.
vertical location of pump compared to
level of influent tank (NPSH).
if want to control flow rate spec and
order flow meter and flow control valve
also.
295
UNEP 2006
Training Agenda: Pumps
Introduction
Type of pumps
Assessment of pumps
Energy efficiency opportunities
296
UNEP 2006
Energy Efficiency Opportunities
1. Selecting the right pump
2. Controlling the flow rate by speed
variation
3. Pumps in parallel to meet varying
demand
4. Eliminating flow control valve
5. Eliminating by-pass control
6. Start/stop control of pump
7. Impeller trimming
297
UNEP 2006
Energy Efficiency Opportunities
1. Selecting the Right Pump
Pump performance curve for centrifugal
pump
BEE India,
2004
298
UNEP 2006
Energy Efficiency Opportunities
1. Selecting the Right Pump
Oversized pump
Requires flow control (throttle valve or by-
pass line)
Provides additional head
System curve shifts to left
Pump efficiency is reduced
Solutions if pump already purchased
VSDs or two-speed drives
Lower RPM
Smaller or trimmed impeller
299
UNEP 2006
Energy Efficiency Opportunities
2. Controlling Flow: speed
variation
Explaining the effect of speed
Affinity laws: relation speed N and
Flow rate Q o N
Head H o N
2
Power P o N
3
Small speed reduction (e.g. ) = large
power reduction (e.g. 1/8)
300
UNEP 2006
Energy Efficiency Opportunities
Variable Speed Drives (VSD)
Speed adjustment over continuous
range
Power consumption also reduced!
Two types
Mechanical: hydraulic clutches, fluid couplings,
adjustable belts and pulleys
Electrical: eddy current clutches, wound-rotor
motor controllers, Variable Frequency Drives
(VFDs)
2. Controlling Flow: speed variation
301
UNEP 2006
Energy Efficiency Opportunities
Benefits of VSDs
Energy savings (not just reduced flow!)
Improved process control
Improved system reliability
Reduced capital and maintenance
costs
Soft starter capability
2. Controlling Flow: speed
variation
302
UNEP 2006
Energy Efficiency Opportunities
3. Parallel Pumps for Varying Demand
Multiple pumps: some turned off during low
demand
Used when static head is >50% of total head
System curve
does not change
Flow rate lower
than sum of
individual
flow rates
(BPMA)
303
UNEP 2006
Energy Efficiency Opportunities
4. Eliminating Flow Control Valve
Closing/opening discharge valve (throttling)
to reduce flow
Head increases:
does not reduce
power use
Vibration and
corrosion: high
maintenance
costs and reduced
pump lifetime
(BPMA)
304
UNEP 2006
Energy Efficiency Opportunities
5. Eliminating By-pass Control
Pump discharge divided into two
flows
One pipeline delivers fluid to destination
Second pipeline returns fluid to the source
Energy wastage because part of fluid
pumped around for no reason
305
UNEP 2006
Energy Efficiency Opportunities
6. Start / Stop Control of Pump
Stop the pump when not needed
Example:
Filling of storage tank
Controllers in tank to start/stop
Suitable if not done too frequently
Method to lower the maximum
demand (pumping at non-peak hours)
306
UNEP 2006
Energy Efficiency Opportunities
7. Impeller Trimming
Changing diameter: change in
velocity
Considerations
Cannot be used with varying flows
No trimming >25% of impeller size
Impeller trimming same on all sides
Changing impeller is better option but more
expensive and not always possible
307
UNEP 2006
Energy Efficiency Opportunities
7. Impeller Trimming
Impeller trimming and centrifugal pump performance
(BEE India,
2004)
308
UNEP 2006
Energy Efficiency Opportunities
Comparing Energy Efficiency
Options
Parameter Change
control valve
Trim impeller VFD
Impeller
diameter
430 mm 375 mm 430 mm
Pump head 71.7 m 42 m 34.5 m
Pump efficiency 75.1% 72.1% 77%
Rate of flow 80 m
3
/hr 80 m
3
/hr 80 m
3
/hr
Power
consumed
23.1 kW 14 kW 11.6 kW
Centrifugal pumps - losses
When assessing the amount of power needed to operate a centrifugal
pump you must always take into account the various losses.
Centrifugal pumps - losses
- Friction loss in bearings and glands, surfaces of impeller
and casing. Some impellers are highly polished to minimize
friction loss.
- Head loss in pumps due to shock at entry and exit to
impeller vanes and eddies formed by vane edges.
- Leakage loss in thrust balance devices, gland sealing and
clearances between cut water and casing and bearing seals.
Centrifugal pumps - losses
A characteristic curve for a centrifugal
pump is obtained by operating the pump at
rated speed with the suction open and the
discharge valve shut.
The discharge valve is then opened in
stages to obtain differentdischarge rates and
total head corresponding to them. The data
can then be represented graphically as a
curve.
The illustration shows the characteristic
curves for three different types of pumps.
Centrifugal pumps - losses
Losses can be caused by:
Failure to deliver
Capacity reduction
Excessive vibration
Failure to deliver caused by loss of suction may be due to
Insufficient supply head
Air leakage at suction pipe (e.g. valve open to empty
bilge, etc)
Loss of priming facility or leaking shaft gland
Centrifugal pumps losses
Capacity reduction could be the result of
A damaged sealing ring
Leaking gland
Obstruction (valve partly closed/foreign body)
Incorrect rotational speed
Excessive vibration may be caused by
Loose coupling
Loose impeller
Bearing damaged
Impeller imbalance
Centrifugal pumps - hydraulic balance
To control the axial movement of the rotating assembly, a balance piston is arranged
to counteract the effect of the thrust of the impellers,especially in the multistage
pumps.
Centrifugal pumps - hydraulic balance
The arrangement keeps the rotating assembly in its correct position
under all conditions of loading .Water at the approximate pressure of
the pump discharge passes from the last stage of the pump between
the impeller hub and the balance restriction bush C into the annular
space B dropping in pressure as it does so .The pressure of water in
the chamber B tends to push the balance piston towards the drive
end.When the thust on the balance piston overcomes the drive and
the impeller thrust the gap A between the piston and balance ring
widens and allows water to escape .This in turn has the effect of
lowering the pressure in chamber B allowing the rotating assembly to
move back towards the pump end .
Theoretically this cycle will be repeated with a smaller movement
each time until the thrust on the balance piston exactly balances the
other axial forces acting on the assembly. In practice the balancing of
the forces is almost instananeous and any axial movement of the
shaft is negligible.
Centrifugal pumps - maintenance
When the pump is due for overhaul, it will
be necessary to dismantle it to its
component parts to examine them for
wear. The following procedures are
intended as a general guide only, and your
attention should be drawn to the
manufacturer's operational instructions
regarding specific pump requirements
before commencing to dismantle the pump.
Centrifugal pumps - shaft sealing
Centrifugal Pumps
Seals
Shaft
Process
Fluid
Leakage
Environment
Vessel
Wall
Centrifugal pumps - shaft sealing
To connect the motor to the impeller, the shaft has to pass
through an aperture in the casing.
To allow the shaft to rotate freely in the casing aperture
there needs to be a gap, but this gap needs to be closed off
to stop air from being drawn in from atmosphere or liquid
from leaking out during operation.
There are two common methods.
Packing
Mechanical seal
The role of the pump, its speed and the type of liquid being
pumped all play a part in deciding which application works
best.
PUMP SEALS
MECHANICAL
SEAL
GLAND PACKING
Stationary type seal
Lantern ring supplies process liquid for sealing ,
cooling and lubrication
Gland packing runs normally on shaft sleeve
Small leakage is allowed and is actually
required to prevent damage to packing &
sleeve
Slight imbalance/eccentricity doesnt effect the
seal function
Can withstand slightly abrasive process liquids
Used for low pressure systems or on suction
sides
Insensitive to installation errors
Wear of shaft or shaft sleeve takes place
Consists of rotating & fixed element
Rotating Seal element is made of Carbon
Fixed seal element is made of Ceramic
Rotating seal is mounted on rotating shaft
and is pushed onto the stationary seal by a
spring
Process liquid is led to the sealing faces for
sealing ,cooling & lubrication
No leakage is allowed through the seal
Slightest imbalance effects the seal function
Abrasives are enemy of mechanical seals
Can withstand high system pressures
Highly sensitive to installation errors
Packing
Packing
A stuffing box with a soft packing material is the traditional
seal for pumps. Normally made from soft impregnated
cotton, which takes the form of a length of square cross-
section wound spirally onto a tube. This enables the correct
length, to suit the external diameter of the shaft, to be
manually cut to the correct size.
The stuffing box is then repeatedly filled with sections until
almost full, the gland can then be tightened down to
provide the axial compressive force. This in turn provides
the necessary radial compressive force required to seal the
gap due to the sloping bottom face of the aperture.
If the force is insufficient the stuffing box will leak, if the
force is too great, the additional friction, and consequently
heat generated by the rotating shaft can damage the soft
packing and/or shaft.
Centrifugal Pumps
Packing
Shaft
Process
Fluid
Leakage
Environment
Vessel
Wall
Centrifugal Pumps
Stuffing box with packing
Centrifugal Pumps
Packing Installation
Centrifugal Pumps
Bypass Flush Line
Bell & Gossett
Series HSC-S
Centrifugal Pumps
Liquid-Abrasives Separator
Centrifugal Pumps
B&G Pump Construction
Standard construction, any impeller
Mechanical seal
Internally flushed, seal cavity
Variety of seal materials
Centrifugal Pumps
Ceramic Seal Insert
Graphite Seal Ring
Compression Ring
Impeller
Retainer (Sec Seal)
Gasket
Mechanical Seal
Rotary Assy
Stationary Assembly
Secondary Seal (Seal Bellows)
Centrifugal Pumps
Seal Cavity
Centrifugal Pumps
Standard construction
Bell & Gossett
Series 1510
Standard
Construction
Centrifugal Pumps
Limits On Seal Performance
Temperature of the flushing water.
225F to 250F
System pH.
7 to 9 for ceramic seals
up to 11 for tungsten carbide seals
Concentration of dissolved solids.
TDS less than 1000 ppm
Concentration of suspended solids.
Less than 20 ppm
Silica, less than 10 ppm
Centrifugal Pumps
Internally Flushed Mechanical Seals
Bell & Gossett
Series HSC
3
Standard
Construction
Centrifugal Pumps
Pump Construction
Stuffing box construction, any impeller
External flush lines
Internal or External Fluid
Compression packing rings
Single flushed mechanical seal
Mechanical seals
Mechanical seals
The provision of rotary shaft seals instead
of the usual stuffing box and gland, where
conditions are suitable, possesses many
advantages. The power absorbed is lower
and is constant, whereas a gland
excessively tightened causes a
considerable increase in power absorbed.
Mechanical seals
In small pumps this may result in overloading the motor. In
addition maintenance costs are reduced, the rotary seal
operating for long periods without wear or attention.
A standard seal consists of a stationary carbon ring insert in
the casing, or seal cover where such is provided, and
against this a metal ring of easy clearance on the shaft
sleeve rotates, contact between the faces being ensured by
a lightly loaded coil spring.
The rubbing faces of both carbon and metal rings are
independently lapped to give a dead flat surface.
A synthetic rubber ring, of circular cross-section, contained
between shaft sleeve and metal ring, in a groove in the
latter, effectively prevents leakage between them.
The diameter of the groove is such that a squeeze is exerted
on the rubber ring, thus a sufficient frictional force is
provided to rotate the metal ring, with certain exceptions.
Mechanical seals
The width of the groove is, however, made considerably
greater so that the metal ring is capable of free axial float with
accompanying rolling action of the rubber ring.
Materials used for the various seal parts are as follows
Carbon stationary ring.
Synthetic rubber ring.
Bronze rotating ring with bronze spring for standard and all
gunmetal pumps.
Stainless steel rotating ring for all iron pumps.
Mechanical seals
For non-lubricating liquids, such as ammonia,
section, contained between shaft sleeve and metal ring, in a
groove in the latter, effectively prevents leakage between
them.
The diameter of the groove is such that a squeeze is exerted
on the rubber ring, thus a sufficient frictional force is provided
to rotate the metal ring, with certain exceptions.
The width of the groove is, however, made considerably
greater so that the metal ring is capable of free axial float with
accompanying rolling action of the rubber ring.
Centrifugal Pumps
Installation
Typical installation
Elbow Support
(Anchor to Base)
Solid Base Foundation
Flow
Pump Motor Hoist Beam
Trolley with Manual
Hook-Type Hoist
Triple Duty Valve
Suction Diffuser
Centrifugal Pumps
Angular Misalignment
Centrifugal Pumps
Parallel or Offset Misalignment
Centrifugal Pumps
Suction Piping Detail
5 dia.
1. Pipe supported
2. Length of suction piping
allows even impeller
loading
RIGHT WRONG
1. Pipe weight hangs on
pump flange.
2. Short suction pipe results
in uneven impeller
loading.
Centrifugal Pumps
Why 5 diameters?
Single Suction
Impeller
Quick quiz
where is the packing material located?
Click on the packing material.
If you are not sure go to the previous
screen to refresh your memory.
Question 1
1: Cavitation of the fluid in a centrifugal pump is caused by?
A) Too high a speed of impeller rotation creating adverse
heat
.
B) The vapour pressure in the suction pipe falls below the
vapour pressure of the liquid at a given temperature.
C) The viscosity of the fluid is too high, the extra power
absorbed being converted into heat.
D) Do not kown.
Question 2
Why is the axial flow pump ideal for trimming and heeling
duties?
A) It is reversible and has a high capacity flow.
B) It has excellent suction lift.
C) Its discharge pressure is increased by the speed of the
ship.
D) I do not know.
Question 3
What is the advantage of a double entry centrifugal pump?
A) It has a lower NPSH required characteristic, giving
advantages in poor suction conditions.
B) It gives double the flow rate.
C) It uses only half the input power for the same flow rate.
D) Don't know
Question 4
Which one of these options is NOT a cause of excessive vibration
in centrifuga pumps?
A) Bearing damaged.
B) Impeller imbaianced.
C) Discharge valve partly closed
D) Don't know
Question 5
Centrifugal pumps need priming because?
A) An excellent suction lift causes the surface of the liquid to
vaporise
B) They must be started with the discharge valve open to
reduce the starting load, but this causes the pump to run
backwards.
C) It is the movement of the liquid from the eye of the
impeller to the discharge that causes a low-pressure region at
the suction, if the pump is started full of air this movement of
liquid does not occur therefore no suction pressure is created.
D) Don't know
Question 6
The energy transformation within a centrifugal pump is as
follows?
A) Pressure is converted into kinetic energy by the impeller;
this kinetic energy is converted to an increase in velocity by
the volute casing.
B) The impeller creates centrifugal force, which increases
the liquid velocity, an increase in velocity means an increase
in kinetic energy, the increased kinetic energy is converted
into pressure by reducing the velocity in the volute casing.
C) Decreasing the velocity in the impeller decreases the
kinetic energy, decreasing the kinetic energy whilst
increasing the velocity of the fluid in the volute casing
increases it's pressure.
D) Don't know
353
Training Session on Energy
Equipment
Pumps & Pumping
Systems
THANK YOU
FOR YOUR ATTENTION
UNEP 2006