S T E A M T U R B I N E M A I N T E N A N C E

Operation
S T E A M T U R B I N E M A I N T E N A N C E

HOW DOES A STEAM TURBINE WORK pressure drop, however, this alone does not account for
all the velocity increase in the steam flow. Combined
A steam turbine is a heat engine, that means it converts
with the pressure drop is a reduction in the heat energy
heat energy into mechanical energy. Other types of heat
in the steam. It is this reduction in heat energy which
engines are gasoline and diesel engines, and gas
accelerates the steam through the nozzle. The
turbines. A steam turbine is NOT a windmill, a
combination of Pressure and Temperature give us a
windmill is not a heat engine, though there are
property known as Enthalpy, this is a measure of the
similarities to a turbine, the primary source of energy is
heat energy available in the steam. Its units ( In the
not heat, it is the velocity of the wind (OK yes, how is
USA) are Btu/lb, i.e. the number of heat energy units in
the wind generated in the first place). A heat engine is
1 lb of steam. Enthalpy is usually given the symbol ‘h’.
one in which a fuel is burned during some part of the
If the enthalpy of the steam entering the nozzle is h 1 and
thermodynamic cycle.
the enthalpy of the steam leaving the nozzle is h 2 then
without going into lots of technical stuff the steam
In the case of the steam turbine the fuel is burned in a
velocity is 224*(H1-H2)1/2 feet/second.
boiler (steam generator), or in some other device, or heat
engine, the exhaust is then used as a heat source to
Now the steam is traveling at some enormous speed,
generate steam.
typically 700 - 1000 feet/second (500-700 miles/hour) or
faster. To achieve these speeds the nozzle passage must
HOW DOES THE TURBINE CONVERT change shape evenly and smoothly. The nozzle also is
HEAT ENERGY INTO MECHANICAL shaped to direct the steam where we want it to go.
ENERGY.
The next item in the steam path is the rotor blade. Its job
is to change the direction of the steam flow, why? Well
The second law of thermodynamics states that heat must a guy called Isaac Newton studied the laws of motion
flow from a high temperature to a low temperature. The and one of the things he found was that a body will
reverse is impossible. In theory it is possible to generate continue to move in a straight line unless an external
steam at high temperature and low pressure and still force acts upon it, therefore to change the direction of
extract energy from it. This is impractical because as the steam flow there has to be a force exerted on the
you heat water, turn it into steam and then raise the steam. This force is the resistance of the blade to move.
temperature even further the expansion is enormous
(typically at atmospheric pressure water expands 1300 So the blade does not want to move, it is attached to the
times when it becomes steam). It therefore makes sense rotor and the rotor is attached to a load. The force of the
to increase the pressure at which you generate the steam blade, on the steam, is at a radius on the rotor. Force x
to reduce the volume. Also by increasing the pressure radius = torque. We all have an appreciation of torque
you can put more heat energy into a given amount of when we tighten, or loosen a nut. Torque x speed =
steam power. If I haven’t lost you yet, it gets just a little more
complicated. Earlier I stated that it was the resistance of
Generating steam at high pressure has another the blade to move which caused a force on the steam to
advantage, it is easier to persuade it to travel along pipes change the steam’s direction. So now the blade moves,
and into the turbine. If there was little or no pressure well another of Mr Newton’s laws of motion is that for
difference between the inlet and exhaust of the steam every action there is an equal and opposite reaction. So
turbine there would have to be some sort of pump to the force the blade exerts on the steam is also the force
move the steam along. This would be very inefficient the steam exerts on the blade. Hence the torque, and
and costly. (Consider a gas turbine, the compressor when the rotor turns we get power (because the blade
absorbs approximately 2/3rds of the energy available has speed).
from the turbine). Generating steam at high pressure is
relatively simple, because you can pump the water into This is all well and good, so why do we not have just
the boiler at high pressure, this is very efficient. one stage in the turbine? The answer is size. There is an
So now we can get the steam into the turbine, once optimum ratio between the velocity of the steam and the
inside it meets with a restriction, this is the first stage speed of the blade. In a pure impulse stage this ratio is
nozzle. The nozzle is specifically designed to convert 0.5 x steam velocity = blade speed. The reason for this
the heat energy, in the steam into velocity. For the will be explained later.
steam to flow through the nozzle there must be a
S T E A M T U R B I N E M A I N T E N A N C E

Firstly we need to know the operating speed of the Now as we have seen above the blades job is to change
turbine. Lets say it is a compressor drive and the the direction of the steam without loss of speed so if the
design speed is 10,000rpm. To limit centrifugal stresses blade turns the steam so that the direction the steam
we have a maximum tip speed, for the blade, of say 600 leaves the blade, relative to the blade, is exactly opposite
ft/s. this gives us a diameter of 600*229/10,000 (229 is to that when it entered the blade, i.e. turned though 180o
a constant when diameter is in inches, speed is in ft/s then the absolute velocity of the steam is Vo -2W
and rotational speed is in rpm) = 13.74 inches. This
leads us to a mean blade diameter of say 12 inches (an
assumed diameter to fit a frame size).

Given this information we can determine the optimum

enthalpy drop for the stage as follows:

Blade speed = 12 *10,000/229 = 524 ft/s

Optimum ratio = 0.5 therefore optimum steam velocity =
524/ 0.5 =1028 ft/s
H1 - H2 = (1028/224)2 = 21 Btu/lb.

From this information we can determine the optimum

pressure at the stage exit. This then gives us all the
information needed to design the stage. 21 Btu/lb
change in enthalpy represents only a small pressure
change, therefore if we need to go to a lower exhaust
pressure we must add stages, using the same calculation,
Now the force on the blade is equal to the mass of steam
until we obtain the correct exhaust pressure at the exit of
the last stage. flowing through it and the total change in relative
velocity through the blade.
If we tried to take all the pressure drop in one stage, with
this size of machine, the efficiency of the stage would be
very low and the stresses on the blade would be very F = w 2 (Vo - W) where w is the mass flow in lb/s.
high. It would also not be possible to design a nozzle to g
physically fit in the machine. The greater the enthalpy
drop across the nozzle the faster the steam velocity, once The power produced is the product of this force and the
the steam velocity exceeds that of sound (i.e. >mach 1) blade speed i.e. P = F x W
the design becomes more difficult and the nozzle
passage length increases.
The available energy from the steam is Pt = wVo2
In some machines use is made of a Curtis stage, or more 2g
correctly a velocity compounded stage. This does allow
a larger enthalpy (or pressure) drop across the stage with Efficiency is P/Pt i.e. what you get/ what you could
reasonable efficiency. The optimum ratio for a Curtis have got! Or:-
stage is around .25 therefore in the above example the
optimum enthalpy drop for the stage is around 95 Btu/lb.
This will result in a steam velocity of 2183 ft/s. Much Efficiency = (w/g) 2 (Vo - W) W
greater than Mach 1 and hence the nozzle design will w ( Vo2/2g)
change considerably.

Why is 0.5 the optimum ratio for an impulse stage? Well, all this reduces rather cleverly to
Let’s assume the velocity of the steam from the nozzle is 4W(Vo - W) = 4 ((W/Vo)-(W/Vo)2)
Vo, this steam should travel such that relative to the V o2
blade it is in the same direction. The velocity seen by
the blade is Vo - W (where W is the blade velocity).
S T E A M T U R B I N E M A I N T E N A N C E

This puts the equation in terms of the blade speed/steam

velocity and if we put various values for this ratio into
the equation we get:-

Ratio Efficency
0 0
.2 .64
.4 .96
.5 1.0
.6 .96
.8 .64 Another loss is that all the steam which passes through
1.0 0 the nozzle does not go through the blade, some hits the
side of the rotor disc and fills the space between the
diaphragm and rotor. This steam passes through the
from this you can easily see that 0.5 is the optimum
pressure balance holes in the rotor disc and hence does
ratio.
no work on the blades. Similarly all the steam leaving
Life is not perfect and you do get losses as the steam
the rotor blade does not go through the next stage
flows from nozzle and in and out of the blade and for an
nozzles. This steam fills the space between the rotor
impulse stage a good maximum efficiency is 80% of the
disc and the next stage diaphragm. This steam passes
above table.
through the interstage diaphragm seal, it is not
accelerated and contributes no work to the stage, the
Over and above this there are other losses which have to
more steam which is allowed to leak between the
be taken into account.
diaphragm and rotor shaft the less there is to be
accelerated, in the nozzle and do work on the blade.
First there is the loss associated with persuading the
steam to flow from the nozzle into the blade which is
When a disc rotates in a fluid medium that medium
hinted at above with the 80% maximum efficiency. The
exerts friction on the disc, this is a further loss of energy.
theory above assumes that the steam leaves the nozzle in
the direction of rotation of the rotor blade. This cannot
happen because the nozzle and blade would have to
share the same space, when this happens we have a
catastrophic failure of the turbine! Because the blade
has to be a distance away from the nozzle there is a
slight axial direction to the steam flow this means that
unless we want the steam to flow upstream when it
leaves the rotor blade we cannot turn the steam through
all of 180o similarly when the steam leaves the blade it
must also have an axial component toward the exhaust
of the machine. This means that we turn the steam
through an even smaller angle. Added to all this the
steam does not travel frictionless through the nozzles
and blades, another loss. This is why we only get about
80% of the theoretical maximum out of the blade.
Worst of all there is an exhaust loss it comprises two
components ‘Leaving loss’ and ‘Hood Loss’. When the
steam leaves the last stage it still has a high velocity, or
Kinetic Energy’ this energy cannot be recovered in a
subsequent stage and is lost to the system. Hood loss is
because of the pressure drop in the exhaust system
which is not used in the turbine.

A loss not often appreciated is that of getting the steam

into the turbine, this is the pressure drop across the Trip
S T E A M T U R B I N E M A I N T E N A N C E

and Throttle valve and the control valves. This is

usually equivalent to a pressure drop of about 5% of the
inlet pressure.

Other losses in the system are more easily understood,

they are the mechanical losses from friction in the
bearings, driving oil pumps, mechanical governors etc.

There are other losses and one will be discussed in the

next section it is associated with the control, or first
stage.

CONTROL
A turbine works at its most efficient when all the nozzle
passages in the turbine are full of steam. During light
load operation, and start up this is not possible. For
optimum efficiency at low loads there should be control OR
valves at each stage, this is impractical and unnecessary.
It is unnecessary because the selection of the steam
turbine is made such that for most of its operating life it
is at or near full load. To assist in starting and control at
low loads in the start-up sequence it is usual, on all but
the smallest machines, to have multiple nozzle control
valves. Each valve only admitting steam to a few
nozzles.

Because of other practical considerations the first stage

nozzles do not usually form a complete circle, this is a
called a partial admission stage. Note that sometimes the
second stage is a partial admission stage, this is to get
blades of a reasonable height and not too small. As the load on the machine increases more steam is
required and other valves open, in turn. They a phased
By only admitting steam to a few nozzles and by such that as one bank of nozzles is passing as much
phasing the opening of each valve it is possible to have steam as it can so the next valve opens. In this way a
many nozzles running full, or nearly full of steam, it is relatively smooth acceleration or increase in steam flow
only in this condition (i.e. full) that the nozzle can is achieved. The aim is to get a straight line relationship
convert the most of the heat energy to kinetic energy. between steam flow and valve lift. This makes deigning
The steam from each control section of nozzles only a governing system much easier. Efficiency also
works on a few blades of the first rotor wheel. This improves as more blades are utilized in the first stage
means that the remaining blades are doing no work at and the later stages are contributing more.
all, in fact they are probably absorbing work and hence
reducing efficiency. PERFORMANCE
From this you can see that during the start sequence and
We have learned what a steam turbine is and some of its
even at low loads it is the first stage , and only part of it
characteristics and why we use different types in
at that, which is doing the work, the contribution from
different applications. Let us now consider how we can
the other stages is small.
estimate the performance of a turbine.

Unfortunately we have to get a little bit technical, we

have to learn how to read a Mollier or steam chart.
Across the bottom is a thing called Entropy, it can never
decrease in value, that’s all you need to know. Up the
S T E A M T U R B I N E M A I N T E N A N C E

side is Enthalpy, this was explained earlier and is the if we divide the number of Btu/lb we are taking out of
heat energy in the steam. The other important lines are the steam, into 2546 we will get the number of lb/.h of
temperature and pressure. Note that pressure is usually steam we need to get 1 horse power. 2546/178 = 14.3
in ABSOLUTE units this means we have to add 14.7 to lb/hp h.
the pressure guage reading ( adding 15 is close enough)
so 600psig = 615 psia. Temperature is read directly in From this we can either estimate the steam flow, if we
degrees Fahrenheit. know the power output of the machine, or estimate the
power out, if we know the steam flow.
At site we are given the inlet steam pressure and
temperature. First we must determine at what point are To estimate the power output divide the steam flow by
these being measured. The closer to the turbine the the steam rate. Lets say the steam flow is 100,000
more accurate will be our estimate. Knowing these we lbs/hour then the power output is 100,000/14.3 or 6993
can plot a point on our diagram. Lets say the pressure is hp. Now we have to take off allowances for mechanical
600 psig and temperature is 750 F. We convert the losses. If there is not a gearbox you could say take off
pressure to absolute units, by adding 15, to give 615 the following:- 1hp to drive the governor, 5 hp to drive
psia. Plotting this in the chart gives and enthalpy of the oil pump, 3 hp for each journal bearing, 30 hp for the
1380 Btu/lb = Hinlet . thrust bearing, i.e. 42 hp this leaves us with 6951 hp
available to drive whatever is connected to the turbine.
If we are looking at a back pressure unit then we need If there is a gearbox allow 2% of the output power.
the exhaust pressure and temperature . Lets say the
To determine steam flow add the parasitic losses above
exhaust pressure is 30 psig = 45 psia, and the to the power absorbed by the driven machine, so lets
temperature is measured at 333 F. First we draw a line assume the same machine, we have 6950 hp being
straight down to where it crosses the 45 psia pressure absorbed by the compressor (measured by a torquemeter
line. The enthalpy is 1126 Btu/lb = H ex(100%), if the or calculated from the compressor particulars), to this
turbine could extract all the energy possible from the we add 42 hp to give 6992 hp, this is the power we need
steam going through it, it would extract H inlet - Hex(100%) or out of the steam. Multiply this by our steam rate of
1380 - 1126 = 254 Btu from every pound of steam. 14.3 lb/hp h we get 100,000 lb /hour of steam required.

We know the exhaust temperature is 333 F so we go

along the 45 psia pressure line until we cross the 333 F
temperature line. ( note because this thing called
entropy cannot decrease you must always go to the
right). We can read off the enthalpy for this condition as
1202 Btu/lb = Hex,actual

This tells us that the turbine only was able to extract

Hinlet - Hex,actual or 1380-1202 = 178 Btu from every pound
of steam going through it. This means the internal, or
thermodynamic efficiency of the turbine is 100*178/254
or 70%. This is the energy we took out of the steam/ the
energy we could have taken out of the steam * 100 to
give percent. Actually this is not too bad for a back
pressure machine. This includes all the throttle losses
through the valves etc
.
Now we have these figures what else can we do with
them, first we can calculate the steam rate, that is the
number of pounds of steam which must pass through the
machine to produce 1 horse power. The magic number
here is 2546 or the number of Btu/hour which is
equivalent to 1 horse power ( or 3412 Btu/h =1kW). So
S T E A M T U R B I N E M A I N T E N A N C E

OK that was the easy part! Now to condensing turbines, This is to get the maximum efficiency out of the LP
well the first thing you should know is that at any given section after steam has been extracted to the process.
pressure water changes to steam at a constant
temperature or vise versa. What has this got to do with One of the quirks of extraction machines is that the
anything? It means that once the steam becomes wet extraction valve is not, it is an induction valve. It works
(that is it contains water droplets as well as steam, it is by allowing the steam not required by the process back
wet steam which you can see) the temperature cannot into the LP section. In fact you can get induction
tell you how wet. This means you cannot determine turbines in which the LP section can flow more steam
where on the Mollier chart the steam should be when it than the HP section. These a rare but the principles are
exits the turbine. In this case you can use the attached the same. Because there is a minimum steam flow
chart to estimate efficiency. We can use the 100% required to keep the turbine cool , a bypass around the
efficiency heat drop multiplied by the efficiency to extraction valve is provided. This ensures that sufficient
determine the steam rate, as we did for the back pressure cooling steam always goes to the LP section.
machine.
The steam consumption curve for a straight back
pressure, or condensing machine is a straight line but for
an extraction (or induction) machine we have to draw a
more complex map, and the more extraction points the
more complex the map becomes. As extraction
machines are the more common we will look at a simple
extraction diagram, this assumes a controlled extraction,
i.e. the pressure at which the steam is extracted is held
constant.

Quality factor is an approximation of the efficiency to

get an idea of the quality factor for a machine in
question divide that theoretical heat drop by the number
of stages, then divide this into the (mean diameter of the
rotor x rpm/229) squared.

Say you have a machine of approximately 32” diameter,

14 stages, running at 3600 rpm with a heat drop of
around 460 Btu/lb. We can calculate (32x 3600/229)2/
(460/14) = 7702, which gives an efficiency of around
77%.

EXTRACTION MACHINES
Extraction machines are really two or more turbines in
series constrained to run at the same speed. They can be
further complicated because the LP section may not First there is the line which represents the performance
have the same mass flow capacity as the HP section. of the machine when there is no extraction, This is
relatively easy to understand and is similar a ‘straight’
S T E A M T U R B I N E M A I N T E N A N C E

turbine performance line. This line ends where the flow

to the LP section is a maximum, this may or may not
coincide with the maximum inlet steam flow. If it does
coincide with the maximum inlet flow then the line will
join directly with the topmost line, which is the
maximum inlet flow. If the maximum LP steam flow is
less than the maximum inlet flow then another line must
join the two. This line represents and increasing inlet
flow with a corresponding increasing extraction flow.

The left most line represents the minimum flow to the

LP sections a minimum flow to the LP section is
required to keep it cool ( we will get onto why later).
The line which joins the latter to the maximum inlet
flow line represents the performance when the
maximum extraction flow is made and the flow to the
LP section is above the minimum required for the LP
section to provide some work.

The intermediate lines represent points of constant

extraction flow.

There is one other line possible, this is a vertical line on

the extreme right of the map. This represents the
maximum power output and is usually shown on
generating sets where there is a means of limiting the
power output.

As you can see this map can become quite complex and
I do not intend to show how it is constructed. It may
explain why the control of extraction turbines is
complex. Today electronic governors have made life
somewhat easier, the electronic extraction governor has
the various corners of the map programmed into it and
the black box does the rest. In the ’good old days’ the
extraction regulation was done by a mixture of levers,
sensing pistons, or bellows, valves and hydraulics.
S T E A M T U R B I N E M A I N T E N A N C E

Impulse Versus Reaction Turbines minimize the generation of thrust from the effect of free-
vortex hub reaction acting on the rotor wheel.

We will now revisit somethings we talked about in the

Stator & Rotor Construction
beginning of this course. First thing I would like to
For the impulse design the stators are diaphragms with
discuss is the difference in reaction and impulse
inner and outer solid rings supporting the stationary
turbines.
vanes, and the impulse rotor is a relatively small
diameter shaft with wheels for each rotating row. The
First of all, all turbines convert thermal energy to diaphragms simply set in grooves inside the case and
mechanical energy through the use of expansion to seal against the case at the OD from the high pressure
create tangential momentum. Yet, although expansion is drop. The rotor penetrates the diaphragm at the ID
a basic process, there are many design choices that where small diameter seals are used to minimize
affect the energy conversion process, one of them being leakage.
the amount of expansion, or energy drop, that occurs
across the moving blade row. The term “reaction” is Seals
used to describe the ratio of rotating blade row energy The pressure drop in the impulse design occurs mainly
drop to overall stage energy drop. If the ratio of energy across the diaphragm. Thus, the rotor shaft diameter is
drop is low then we call this turbine and impulse turbine. kept as small as possible consistent with rotordynamic
If this energy drop is high we call this a reaction turbine. criterion and seals are employed at the diaphragm
penetration to minimize leakage. To further improve
Impulse Turbines efficiency, advanced seals have been developed for the
The impulse turbine design refers to a low reaction stage rotor blade hub and tip that mitigate the effect of free-
design, which has all, or most, of the energy drop vortex reaction and disc windage. The combination of
occurring across the stationary vanes. hub seal and pressure equalization holes ensures that the
diaphragm leakage does not re-enter the mainstream
flow and disrupt the high tangential momentum flow.
Characteristics Tip seals are used to reduce leakage and mixing losses
associated with the free-vortex reaction at the tip.
Stages
The impulse stage design achieves peak efficiency at a Reaction Turbine
lower velocity ratio (W/Vis) than the reaction design. The reaction turbine design refers to a high reaction
Therefore, the impulse design requires fewer stages to stage design, which has equal energy drop across the
achieve the same energy drop for a given wheel speed. stationary vanes and rotating blades.

Airfoils Characteristics
Impulse turbine airfoils are necessarily different for the
stationary and rotating rows because of the differing
velocity triangles. The greater work per stage of the Airfoils
impulse design results in a rotating airfoil that is larger The velocity triangles for a reaction design are similar
and stronger than the reaction design due to the greater for the stationary and rotating rows. Therefore, the same
turning of the profile and the greater width. airfoil profile can be used in both rows. In addition,
Furthermore, the greater width of the impulse blade because of the decreased work per stage, there is less
results in fewer blades per row and a larger throat turning in a reaction airfoil, and with less bending stress
opening, which is therefore less susceptible to foaling the profile is made very narrow. This lower turning and
from steam deposits. increased aspect ratio (height-to-width ratio) results in a
slightly more efficient airfoil profile. However, the fact
that the reaction airfoil is smaller also means that the
Rotor Thrust
throat width is smaller and steam deposits seriously
With little or no pressure drop across the impulse rotor
degrade their performance.
blade there is little or no thrust on the rotor. Pressure
equalization holes in the wheels are also utilized to
S T E A M T U R B I N E M A I N T E N A N C E

Rotor Thrust Bearings

In a reaction turbine design there is considerable thrust Even though a balance drum is used in the reaction turbine
generated due to the imposed pressure drop across the the high thrust load of this design combined with the
rotating blades. Because the rotor thrust is so large a potential for flowpath deposits and balance drum seal
balance drum is typically used to compensate for the deterioration requires that a large thrust bearing be used for
load. Good sealing of the balance drum must be reliability. Also, larger journal bearings are required to
maintained to achieve good efficiency and maintain accommodate the greater weight of the reaction drum rotor.
thrust load. Small reductions in sealing effectiveness
and blade throat opening can have a significant impact
on the thrust.

Stator & Rotor Construction

The reaction turbine design rotor is a “drum” style,
which has a very large base diameter from inlet to
exhaust. This is done to utilize the same airfoils in both
the stator and rotor. The stator vanes are either installed
directly into the case or are carried in thin rings. Many
small, large diameter seals are used at the stator ID and
the rotor OD to minimize leakage from the imposed
pressure drop.

A reaction turbine may also employ an impulse stage to

quickly reduce the inlet pressure in order to minimize
case bolting load when inlet pressure is high. At even
higher inlet pressure the impulse turbine can employ a
velocity-compounded stage (i.e curtis) to quickly reduce
pressure loads on the case.

These different construction features also affect

operational characteristics. The drum rotor of the
reaction design has significant thermal inertia; and
therefore load transients common in mechanical-drive
service must be conducted with extreme care to prevent
rubbing the reaction seals. The impulse rotor does not
have as much thermal inertia and is able to
accommodate more rapid load changes.

Seals
In the reaction turbine many small seals of large diameter
are typically caulked into the rotor and stator to limit
leakage due to the pressure drop across the stator and rotor.
Because of the drum rotor the seals have a large diameter,
which means that the seals have a large leakage area as they
open up due to wear and damage. Moreover, the small size
of the reaction seals makes them susceptible to rapid
deterioration from rotor transients, handling, steam deposits
and FOD.
S T E A M T U R B I N E M A I N T E N A N C E

Commissioning Process closed and the ESV / Trip valve is open, the rotor
should not turn. This will allow steam into the steam
chest and warm the steam chest. Once steam chest
After the mechanical erection of the turbine or the major temperature is normal and all other prerequisites are
overhaul of the turbine, there needs to be a process of met, we can start the turbine.
commissioning or recommissioning the turbine. This
will involve checking all the instrumentation loops to
Typically you would want to slow roll the turbine
verify correct wiring and calibration. The checks of all
around 500-1000 rpm for 30 minutes to warm the
the auxiliary systems for proper functions and the
turbine. This will vary from turbine to turbine. The tech
operational check of the control system. As a minimum
manual should have a start up curve. Review the client’s
all safety device must be calibrated and checked. These
startup procedure to insure it complies with the OEM
checks should start with the turbine in a cold safe
startup instruction. Do not rely strictly on the client /
condition and test inputs to the various devices to verify
operator’s startup procedure or their startup experience,
operation.
typically they are wrong. The next step up in speed will
be to approximately 3000 rpm for 30minutes depending
Turbine Startup (General) on the 1st critical speed. AVOID ANY OPERATION IN
CRITICAL SPEED BANDS.
We will look at the typical startup of a condensing
turbine since this would be the most common type we After this step the speed can be raised to minimum
see here at Sulzer. governor speed. At minimum governor speed, the
turbine should be allowed to stabilize, this means that all
temperatures, pressures, flows and any other parameters
The Start Up of the steam turbine starts with the
warming of the steam piping. The Main Steam piping should not be changing. This should take approximately
30 minutes.
will have bypass valves around the main valves and
drains. The valves should be cracked open to allow the
steam to slowly warm the piping. The auxiliary steam At this point on a normal startup we would be ready to
piping to the air ejectors will also have to be warmed to load the turbine. Since the turbine can be driving
normal operating conditions. If the valves are opened generators, compressors or pumps, the loading will be
too much or too fast, the condensate will flash to steam different from machine to machine. The loading should
and cause “Water Hammer” in the piping. This can take place gradually over time and the turbine allowed to
damage piping and equipment. Warm Up is a slow stabilize at each step.
controlled process to avoid this condition.
If we have overhauled the turbine, there should be some
If the Turbine has a turning gear, verify all prerequisites minimum checks prior to loading the turbine. The first
are met and place the turbine on the turning gear. Not all thing we would do after reaching minimum governor
turbines have turning gears so the start up procedure will and stabilizing the turbine would be to slowly raise the
vary. speed to maximum governor speed. This is the
maximum speed that the turbine will normally operate.
The gland exhaust condenser and air ejector condenser We should raise speed gradually to maximum governor
and record all parameters. At this point if we are
should be placed on service. Once this is done, start
drawing a vacuum in the main condenser. At this time required to perform an actual overspeed test, we would
raise the speed to the overspeed setpoint ±10%. Do not
you can commission the gland seal steam if you have a
turning gear. The shaft has to be rolling to start gland stay at any speed over the maximum governor speed for
any length of time. The turbine is under higher than
seal steam due to possible bowing of the shaft.
normal stresses at this point and time should be limited.
Once the overspeed is complete, restart the turbine and
Once the piping has been warmed and the inlet steam up
bring to minimum governor. We can now load the
to the ESV/Trip valve is at normal operating pressure
turbine.
and temperature and all turbine start prerequisites are
met, Reset the Turbine. On some turbines this will open
the ESV / Trip Valve on others you will have to open
the trip valve separately. If the governor valves are full
S T E A M T U R B I N E M A I N T E N A N C E

When we load the turbine, the load should be added

gradually. The firs step should be around 5% load and
hold until parameters stabilize. Then load should be
increased in 5% – 10% increments stabilizing at each
point until maximum load is reached.

Normal Shutdown (General)

The normal shutdown procedure would be the reverse of
the startup procedure.

The turbine load should be gradually reduced until the

turbine is completely unloaded. The turbine speed
should be lowered to minimum governor speed and the
turbine will be tripped and allowed to coast down.
Insure the turning is engaged once the turbine stops
turning to prevent rotor bow. Some turbines do not have
a turning gear. In this case the turbine will roll to a stop.
The sealing steam must be isolated before the turbine
stops turning to prevent rotor bowing. The vacuum
system would be secured. The lube oil system would
remain running until the bearing temperatures are the
same as lube oil supply temperature and the turbine
interior temperature is less than 100C. The turning gear
should not be stopped until the turbine internal
temperature is less than 100C.
S T E A M T U R B I N E M A I N T E N A N C E

TYPICAL TURBINE STARTUP PROCEDURE

TYPE OF START COLD START WARM START HOT START HOT RESTART

UNIT STATUS  

Hours Since More than 24 hrs. Less than 24 hrs. Less than 8 hrs. Less than 30 min
Shutdown

UNIT CONDITION  

A. Turning Gear One hour minimum Continuous Continuous Continuous

(with seal steam) (1 hour minimum) (1 hour minimum)

Rotor eccentricity less Rotor eccentricity less Rotor eccentricity less Rotor eccentricity less
than 1.5 mils, casing than 1.5 mils, casing than 1.5 mils, casing than 1.5 mils, casing
differential expansion differential expansion differential expansion differential expansion
less than 0.165". less than 0.165". less than 0.165". less than 0.165".

B. Rollover 200 - 250 RPM for 1 200 - 250 RPM for 1 200 - 250 RPM for 1 200 - 250 RPM for 1
min. If vibr. Level is min. If vibr. Level is min. If vibr. Level is min. If vibr. Level is
less that1.5 mils, less that1.5 mils, less that1.5 mils, less that1.5 mils,
proceed to 500 RPM proceed to 500 RPM proceed. proceed.
and hold for 10 and hold for 5
minutes. minutes.

C. Slow Roll Raise at 100 Raise at 150 Raise at 200 Raise at 200 RPM/min.
RPM/min. to 950- RPM/min. to 950- RPM/min. to 950- to 950-1000 RPM hold
1000 RPM hold for 30 1000 RPM hold for 15 1000 RPM hold for 10 for 2 minutes
minutes minutes minutes

D. Raise to Minimum Raise at 100 Raise at 150 Raise at 200 Raise at 400 RPM/min.
Governor. (Avoid RPM/min. RPM/min. RPM/min.
dwelling on turbine
rotor first response).

E. Minimum Hold Minimum Hold Minimum Hold Minimum Hold Minimum

Governor hold governor speed for 10 governor speed for 10 governor speed for 5 governor speed for 2
minutes. minutes. minutes. minutes.

F. Increase speed to Increase speed to Increase speed to Increase speed to Increase speed to 100%
100% & synchronize 100% and synchronize 100% and synchronize 100% and synchronize and synchronize

G. Place generator at Hold at 10% load for Hold at 10% load for Hold at 10% load for Hold at 10% load for 5
10% load 10 minutes 5 minutes 5 minutes minutes

H. Increase to 100% As required As required As required As required

load. (Monitor
supervisory systems)

I. Approximate time 80 minutes 55 minutes 30 minutes 15 minutes

to synchronization