8.38
Steam Turbine Controls
STM T
E. J. FARMER
(1985)
B. G. LIPTÁK
(1995, 2005)
Flow sheet symbol
Steam turbines are used in the generation of electric power, particularly in combined cycle power stations (Section 8.33), and also as variable- or constant-speed drives (Section 7.10) for larger rotary equipment, such as compressors and pumps (Sections 8.15 and 8.34). Their controls are often integrated with the steam generator (Section 8.6), which can use fossil or nuclear fuel. In the first part of this section, the nature and the operating principles of steam turbines will be discussed first. This will be followed by a description of the various types of turbine designs, efficiencies, and applications. In the second part of this section, the control of steam turbines will be covered. This discussion will be started with the description of the traditional steam governors and basic regulatory turbine controls. The controls of pressure let-down and extraction turbine applications and the methods of eliminating interaction and decoupling will also be discussed. The section will also describe the steam turbine safety systems and will be concluded with a description of optimization and advanced controls. The optimization strategies described will consider the nature of the installation, which might have the goal of maximized electric power production while meeting the steam demand of the plant or the goal of maximizing direct steam utilization. The discussion of advanced controls will cover the use of model-based predictive control and self-diagnostics for maintenance purposes. It will also touch upon thermal stress monitoring (TSM), turbine protection (TP), and monitoring/ sequential control (MSC).
Such variable-speed operation is an advantage, if the turbine is used to drive pumps or compressors. Figure 8.38a shows the relative operating ranges of different types of compressor drives. When provided with the appropriate speed governor, steam turbines can provide excellent speed stability, which is desirable when the turbine serves as the prime mover in electric generators. In comparison to all other prime movers, steam turbines are very reliable. Their availability factors are high, and their maintenance costs are low. This is largely a result of the inherently balanced design that is completely free of
106
Compressor inlet flow, ft3/min
INTRODUCTION
105
Steam turbine
Gas turbine 104
103
Electric motor, gear driven
2
4
6 8 10 12 14 16 18 20 22 24 26 28 Compressor speed, thousand RPM
Speed governor classification
Characteristics Steam turbines are energy conversion machines. They extract energy from the steam and convert it to work, which rotates the shaft of the turbine. Steam turbine sizes range from shaft output energies of a few kilowatts to well over 1000 megawatts, and there is no reason why still larger machines could not be built. No other prime mover can achieve the shaft output capability that is easily attained by large steam turbines. Rotational speeds vary from approximately 1,800 to 14,000 rpm; this speed can be modulated over a wide range.
Regulation
Variation
A
10%
¾%
B
6%
½%
C
4%
¼%
D
½%
¼%
Governor class
FIG. 8.38a The throughput of a compressor with a steam turbine drive is higher than those of gas turbine or electric motor-driven units.
2137 © 2006 by Béla Lipták
2138
Control and Optimization of Unit Operations
Entropy, (B.T.U. per LB. deg. fahr.) 1.0 1650
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
2.1
2.2
2.3 1650
ture deg. fahr, 1200 10 Constant tempera 00
1600
1600
1100 90 0
1550
1550
1000 80 0
1500
1500
900
70
0
1450
800
60
1450
0
1400
700
50
r. 10
100
1150
1100 ∆h = 325 4−5
so .p
er
h4−5 = 73%
LB re ,
re s
su
P2 15
P5
1000
Co
ns
ta
1000
1050
nt p
∆h1−3 = 325
.0
5
1200
0.2
10
200
0.5
1050
300
0
line
∆h1−2 = 185 h1−2 = 67%
0
.1
1100
erc ent
400 1250
20
BS
re p
0
.A
mo istu
Sat ura tion
1350
Enthalpy, B.T.U. per LB.
Sta n
Fah
1400
500 1300
30
5
ant
eg .
0
.in
1150
,d
14 20 .69 6 10
Cons t
pe rhe at
600
40
2.5
P1
0
1200
100
1250
sta nt su
500 300 200
400 0 300 200 0 0 150 1000 0
Con
50 30
1300
dar da tm o sp
P4
550
Enthalpy, B.T.U per LB.
1350
her e
0
950
950
∆h4−5 = 445
20
900
25
900 P3
P6
850
850
30
40
800
35
800 50
750 1.0
1.1
1.2
1.3
1.4
1.5 1.6 1.7 1.8 Entropy, (B.T.U. per LB. deg. fahr.)
1.9
2.0
2.1
2.2
750 2.3
FIG. 8.38b Mollier diagram showing performance of a steam turbine.
reciprocating or rubbing components (except, of course, for the bearings).
OPERATING PRINCIPLES The amount of energy that the steam turbine extracts from the steam depends on the enthalpy drop across the machine. The enthalpy of the steam is a function of its temperature and pressure. Because the operating conditions of steam turbines are generally known in terms of inlet and outlet temperature
© 2006 by Béla Lipták
and pressure, one can use a Mollier diagram as a graphic tool to determine the amount of energy available under a particular set of conditions (Figure 8.38b). As an example, one might consider the case where the turbine inlet conditions correspond to point P1 and the outlet conditions to point P2 in Figure 8.38b. A line drawn between these two points is called the “expansion line” and represents the operation of the turbine as it is extracting energy from the steam. In actual operation, this line is not straight, and its shape does depend on the internal operation of the turbine.
8.38 Steam Turbine Controls
In an ideal turbine, the steam would expand at a constant entropy (isentropically). The condition of the exhaust steam from an ideal machine that has no losses would correspond to point P3 on Figure 8.38b. This P3 point is found by drawing a vertical line from the inlet point P1, down to the same exhaust pressure as the actual machine is operating at. Steam Turbine Efficiency If one refers to the change in enthalpy from P1 to P2 as ∆h1−2 and to the change in enthalpy between P1 and P3 as ∆h1−3 , one can use these quantities to calculate the turbine’s efficiency and steam rate. The efficiency of the turbine, neglecting mechanical losses, is found by
η=
∆h1 − 2 ∆h1 − 3
8.38(1)
The steam rate that an ideal machine would require to operate as theoretically predicted is called the theoretical steam rate, QT , and may be found as QT =
2545 lbm ∆h1 − 3 HP − hr
8.38(2)
The actual steam rate, QA, is calculated by dividing the theoretical steam rate with the efficiency, hence, QA = QT /η
lbm HP − hr
8.38(3)
This somewhat simplified perspective of steam turbine performance can help the control engineer in both understanding and optimizing steam turbines. In order to operate the turbine at maximum efficiency, the steam should leave the nozzles and impact on the turbine blades at sonic velocity. In order to maintain the nozzle jets at sonic velocity at partial loads, it is necessary to shut off some blocks of nozzles so that the active ones will receive approximately the same amount of steam all the time. The efficiency of steam turbines increases with size, with the superheat temperature of the steam, and with the level of vacuum at the turbine exhaust. Example As shown in Figure 8.38b, if saturated steam enters the turbine at 300 psia and is exhausted at 2 psia, then from each pound of steam, 325 BTUs of energy can theoretically be recovered in an “ideal” turbine. If the actual expansion line is P1 − P2, the actual energy recovery is 185 BTUs per pound of steam, and therefore the efficiency of the installation is 67%. In actual practice, turbines are not supplied with saturated steam. This is because condensation could cause turbine
© 2006 by Béla Lipták
2139
blade erosion due to the presence of water droplets at the high jet velocities, and also because much more energy can be recovered if the steam supply is superheated and if the exhaust pressure is reduced. If, for example, the same 300 psia steam enters the turbine with 200°F superheat and is exhausted at 0.5 psia (instead of 2 psia), the theoretical energy recovery will rise to 445 BTUs per pound (P4 − P6). Under these operating conditions, the actual energy recovery will rise to 325 (from 185) BTUs per pound. This obviously represents a major improvement in the performance of the machine. This improvement can also be expressed in terms of steaming rate. The steam rate of the turbine is the amount of steam needed to provide one horsepower-hour (hp-hr) of work, which is the equivalent of 2545 BTUs. If the actual operation of the steam turbine corresponds to P1 − P2 on Figure 8.38b, the steaming rate is 13.8 lb of steam per horsepower-hour, while if it corresponds to P4 − P5 it is only 7.8 lb of steam per horsepower-hour. Advantages and Limitations Even though the initial cost of steam turbines is usually higher than that of alternative prime movers, there are some benefits that can mitigate this cost difference. Especially in larger sizes, steam turbines are physically smaller than most other prime movers, consequently, they require less floor space. This can decrease the cost of the building in which they are to be installed. Because the inherently balanced design of steam turbines produces considerably less vibration than do reciprocating machines, equipment foundations can also be considerably lighter. Because steam turbines are less likely to cause fires than are other prime movers, they also have an advantage in applications in locations where flammable materials are present. In applications where the prime mover has to endure substantial overloads, the steam turbine also has an advantage. It can tolerate, without damage, such overloads that would severely shorten the service life of alternative prime movers, if they could tolerate them at all. The optimization of steam turbine systems is well understood and more widely practiced than the power consumption optimization of alternatives such as electric motors and gas turbines. Less energy will usually be required for a steam turbine prime mover to drive a load than would be necessary if the size of a turbine driving an electrical generator were increased to provide the electric power to drive the same load with an electric motor. Besides initial cost, the major disadvantage of steam turbines is their low tolerance for wet or contaminated steam. Wet steam can cause rapid erosion, and contaminants can cause fouling. Both will reduce the turbine’s efficiency and will shorten its life. Steam quality monitoring is therefore an important requirement to maintain the reliability and to reduce the operating cost of steam turbines.
2140
Control and Optimization of Unit Operations
TABLE 8.38c Performance Data for Condensing and Noncondensing Steam Turbines Technical Data
Units
Output
MW
rpm
min
Condensation Turbine
-1
Back-Pressure Turbine
5–120
5–120
Up to 14,000
Up to 14,000
Inlet pressure
bar (psi)
Up to 130 (1,885)
Up to 130 (1,885)
Inlet temperature
°C (ºF)
Up to 570 (1,058)
Up to 570 (1,058)
Discharge pressure
bar (psi)
Up to min. 0.02 (0.29)
Up to 40 (580)
Exhaust flow
M /s (ft /s)
3
3
Up to 1,300 (45,930)
TURBINE DESIGN CONFIGURATIONS The energy source for steam turbines is the pressure difference between the supply and exhaust steam. The higher this pressure difference and the higher the superheat of the steam, the more work the turbine can do. The two main categories are the condensing and the backpressure turbines (Figure 8.38d). Table 8.38c provides a range of operating conditions and electrical power production rates for both types of steam turbines. The exhaust pressure of a “condensing” turbine is usually subatmospheric, while that of a “noncondensing” or “backpressure” turbine is greater than atmospheric. Condensing turbines are most often used for electric power generation, while back-pressure turbines are utilized in cogeneration power plants (Section 8.33), which simultaneously supply steam and electricity for the users. Steam turbine installations can also be configured with more than one exhaust steam streams. A second, and sometimes
—
even a third, outlet may be provided to allow the “extraction” of steam at different pressures (Figure 8.38e). A rare combination of a mixed input and an extraction turbine is one in which steam can be removed when load conditions permit, or steam can be inducted when additional energy is necessary. This unit is usually referred to as induction– extraction turbine. Application Configurations A back-pressure turbine with its inlet connected to the plant’s high-pressure header and its outlet supplying steam to an intermediate header is called a topping turbine, because it is
High pressure SC
High pressure
Stage 1 SC
SC
Condensing
Stage 2
Noncondensing (backpressure)
Intermediate pressure
Intermediate pressure
PC
Condensate Condensate
FIG. 8.38d Typical installation of condensing- and backpressure-type steam turbines.
© 2006 by Béla Lipták
FIG. 8.38e The extraction turbine generates as much intermediate pressure steam as required by the plant, while taking as much high-pressure steam from its supply header to meet the load by maintaining its shaft speed constant.
8.38 Steam Turbine Controls
Steam
HPS
HPS
HPS T
2141
T
1st
2nd
PC
Exhaust of other turbine T
MPS PC Condensate Condensing
Non-condensing (back-pressure)
T
Condensate
Extraction
MPS
Condensate Topping
Bottoming
FIG. 8.38f Terminology used to describe various steam turbine installations.
using steam from the “top” of the plant’s thermodynamic cycle. Similarly, a turbine installed between the noncondensed exhaust of another turbine and the condensate system of the plant is called a bottoming turbine. A topping turbine could also be described as a noncondensing or back-pressure unit, and a bottoming turbine as merely a condensing turbine. Figure 8.38f describes the main variations of turbine configurations. The terms “topping” and “bottoming” as generally used describe turbines that are designed to work over unusual ranges of pressures. For example, a bottoming turbine is unusual in that it is generally designed for operating at a very low inlet pressure, and hence the term “bottoming” is more descriptive than the more generic term of “condensing.” Steam turbines can also be classified according to the purpose of their installation. If a turbine’s purpose is to generate electricity, it might be called a “generator drive,” while if its purpose is to drive a pump or compressor, it can be called a “mechanical drive.” Internal Design Configurations From an internal design perspective, the steam turbine is either an “impulse”- or a “reaction”-type design. In the United States, almost all turbine designs are of the axial flow variety, and only a small number are the tangential flow variety. In Europe, a significant number of turbines are the radial flow design. The steam turbine can also be single-stage or multistage; if multistage and as a function of the number of parallel exhaust stages, one would refer to the turbines as single-flow, doubleflow, and so forth. The casing and shaft arrangement is also an important way of categorizing turbines. In a single casing machine, there is a single casing and one shaft. In a “tandem” design, there are two or more casings connected end to end by a shaft. In “cross-compound” configurations, there are two or more casings connected by multiple shafts. As can be seen from the above, a turbine can be described on the basis of at least three methods of classification.
© 2006 by Béla Lipták
STEAM TURBINE GOVERNORS There are two broad categories of steam turbine controls: “safety systems” and “process systems.” Both will be described in the paragraphs that follow. Safety systems are intended to eliminate or, at least, to minimize the possibility of damage to the machine or the hazard to operators. Process control systems serve to control the operation of the machine, so as to follow the load in a stable and efficient manner. The Governor Valve One valve, the governor valve, is common to all turbine applications. This is the valve between the main steam supply and the turbine. This valve is the primary means of controlling the unit. When the demand for energy from the turbine is changed, it is the opening of this valve that changes to match the new demand by introducing a new supply of steam energy. The energy supply and demand is matched when the turbine speed is constant. If the supply valve is too far open, the turbine will run at a speed above that desired. If the valve is too far closed, the turbine will slow down. In essence, the governor valve controls the flow of steam, generally measured in pounds per hour, into the turbine with the assumption that inlet and outlet conditions are constant. When that is the case and the shaft speed is constant, there is a balance between the steam flow and the shaft horsepower. Some of the steam turbines can only be operated at constant speed because the characteristics of their steam supply valves (the governor valves) are not suited for throttling, because of their quick-opening characteristics. A quick-opening valve plug (Figures 6.19a and 6.19f) is like the plug of a bathtub in which a slight lift results in nearly maximum flow. By changing that characteristic, a constant-speed turbine can be changed into a variable-speed one. Figure 8.38g illustrates how, by welding into the governor seat a characterizing ring, the initially quick-opening characteristic can be altered and the turbine can be changed to operate at variable speed.
2142
Control and Optimization of Unit Operations
with a hinge in the center to allow vertical motion of each. By gearing to the shaft of the turbine, the assembly was made to rotate. As the shaft speed increased, the weights lifted up toward horizontal. A linkage controlled the throttling valve by admitting less steam as the weights rose and more as they fell. This system was the beginning of automatic machine control and is still used almost unchanged in some modern mechanical governors.
B
C
Part B Part A Quick opening plug A Part A
Part B A + 0.03"
0.2"
Hydraulic Droop Governors
B + 0.03"
30° 0.1"
C + 0.1" 0.2"
30° C + 0.5"
0.5"
FIG. 8.38g A double-seated steam governor valve can be rebuilt for optimized variable-speed service. The notched rings provide the necessary rangeability.
The Early Speed Governors The speed of the turbine is controlled by the “governor.” Governors can be mechanical, hydraulic, and electrical. They all include a pilot valve, or a more sophisticated controller, which modulates the turbine’s inlet valve in order to keep the shaft speed on set point. Mechanical governors have been developed from James Watt’s original flyball governor shown in Figure 8.38h. The assembly consisted of two weights on the end of short arms,
Steam supply
FIG. 8.38h Schematic representation of flyball governor. As shaft changes speed, rotating fly balls move up and down. Linkage then controls steam supply valve to regulate steam rate. If shaft speed is too fast, balls move up, raising linkage, which then closes down the inlet valve. If shaft speed is too slow, balls drop, lowering collar and opening the inlet valve.
© 2006 by Béla Lipták
In hydraulic governors, the shaft speed is generally detected by a flyball, but instead of a direct mechanical linkage between the position of the flyballs and the control valve, a hydraulic system is used to amplify the force generated by the flyball position. The amplification feature improves control sensitivity, because very small changes in flyball position, corresponding to small changes in shaft speed, are sufficient to produce effective control actions. In addition to amplification, the signal can also be characterized as needed for stability. Two types of governors are distinguished: In the “isochronous” design, the objective is to maintain the speed of shaft rotation constant regardless of load, while in the “droop” design, the speed of the machine is deliberately decreased as load increases (Figure 8.38i). The droop governor is a proportional-only controller, which cannot change its output control signal without first developing an error. This “offset” phenomena was explained in connection with Figure 2.2e in Chapter 2. The terms “offset” and “droop” are interchangeable. On an increase in speed, the flyballs of a droop governor move out, which raises the stem on the pilot valve. This movement is opposed by the spring. Pressure is applied from the pilot valve to the top of the actuator; the bottom of the actuator drains to the sump through the pilot valve. This decreases the throttle setting. As the linkage moves down, the spring force increases and the force provided by the flyballs is exactly opposed. This moves the pilot valve back to the null position, which maintains the new lower speed. Droop governors can be an advantage in some applications. For example, if two steam turbines are used to drive two electric generators that are electrically connected in parallel, the droop characteristic will allow the generators to share the total load, whereas isochronous governors would not. If both generators were to run at exactly the same speed, the division of load between them would depend only on the electrical characteristics of the generators. If they were not precisely identical, which is always the case, the division of load between them would be unequal. If the speeds were not perfectly matched, one might carry all of the load. A droop governor, however, would cause the one turbine with the heavier load to tend to slow down. When the load carried by the first generator matched that of the other unit and began to surpass it, the first unit would slow down. An
8.38 Steam Turbine Controls
2143
Increasing Throttle setting Decreasing
Actuator
Pressure
Pressure relief lines Sump
Pilot valve
FIG. 8.38i The design of a droop governor.
equilibrium would be quickly established with each unit carrying a share of the load. The problems requiring compensation generally include instability that occurs during speed changes. In high-gain control systems, buffering is used to minimize instability. In this case, the governor introduces droop on all speed changes and controls the rate at which the temporary droop characteristic is removed. In this way, speed transitions can smoothly be made. The droop due to buffering can be built into a governor whether the device is a droop type or not. Electronic Governors Electronic governors perform the same functions as their mechanical-hydraulic counterparts, but in a somewhat different way. The flyballs are replaced by an electronic tachometer input that is usually generated by a magnetic sensor. The sensor can be triggered when the teeth of a gear connected to the machine’s shaft pass by it (Figure 8.38j). The varying reluctance of the magnetic circuit is used to generate a periodic function with a frequency proportional to the rotational speed of the shaft. The control valve is most often throttled hydraulically, although its actuator can also
Permanent magnet
Output signal
Sensor winding
Rotation
Toothed wheel (magnetic material) driven by rotating member
FIG. 8.38j Schematic of an induction-type speed sensor.
© 2006 by Béla Lipták
be electronic or pneumatic. The characteristics of the controller are determined by its transfer function, which usually provides similar performance as that of the earlier designs. The main difference is that a wider range of features can be built into a single unit, and the same design can be easily adapted to a variety of applications. For example, in addition to speed control, the governor can maintain either the inlet or the exhaust pressure or can control or manipulate other process conditions. Further, it can automatically parallel generator sets, provide overspeed protection, and monitor other machine safety devices so a shutdown can be effected in the event of an unsafe condition. Electronic governors provide versatility rather than improved performance. Advanced Governors Advanced control systems perform speed control, load control, steam pressure control, valve testing, remote control, and turbine protection. Normal operation, in addition to speed, load, and steam pressure modulating control, also includes valve testing and remote control (autodispatch, autosynchronizer, and so on). Even the advanced controls usually operate the high-pressure and low-pressure valves through the existing electrohydraulic controls of the turbine. Further information on steam turbine performance and speed control can be found in API Standard 611, “GeneralPurpose Steam Turbines for Refinery Services,” API Standard 612, “Special-Purpose Steam Turbines for Refinery Services,” NEMA Standard SM21, “Multistage Steam Turbines for Mechanical Drive Service” (Table 8.38k), and NEMA Standard SM22, “Single-Stage Turbines for Mechanical Drive Systems.” Performance objectives of governors are covered in NEMA Standard SM22-3.13. Governor systems are classified as A, B, C, or D, depending on performance objectives. Table 8.38k summarizes the basis for these classifications.
2144
Control and Optimization of Unit Operations
TABLE 8.38k Governor Classification and Performance per NEMA SM21 Speed, a Range, %
Class
Maximum Speed b Regulation, %
Maximum Speed c Variation, ±%
Maximum Speed d Rise, %
Trip Speed e Setting, %
10
0.75
13
115
A
10, 20, 30, 50, 65
B
10, 20, 30, 50, 65, 80
6
0.50
7
110
C
10, 20, 30, 50, 65, 80
4
0.25
7
110
D
10, 30, 50, 65, 80, 85, 90
0.5
0.25
7
110
a
Governor may be adjusted to produce any speed within this percentage of rate speed. Maximum speed regulation from no load to full load. c Maximum speed variation when operation is at constant load. d Maximum overspeed that can occur under any operating conditions. e Proper overspeed trip setting to coordinate with governor maximum speed rise. b
CONTROLS AND OPTIMIZATION A number of control systems will be described here in an order of increasing sophistication. These will include the controls for pressure let-down and extraction turbine controls. Controls systems will also be described for the decoupling of interaction between control loops and for optimization purposes.
ernor will detect this increase in speed and act to eliminate it. Its means of doing so is to reduce the energy supplied to the turbine by closing the supply valve. If the net change in the energy balance were negative, the shaft would slow down and the governor would respond by opening the supply valve. Pressure Let-Down Control
The Basic Turbine Controller The simplest application is one in which a turbine is used to operate a mechanical load at constant speed. Here, steam is supplied from a header and is condensed in the turbine (Figure 8.38l). In this case, the speed controller (governor) senses the shaft speed and manipulates the steam supply valve to keep the speed on set point. Variations in load, caused by either shaft loading or variations in supply header pressure, affect the balance between the energy supplied to the turbine from the steam system and the work removed from the turbine’s shaft. If more energy is available than is being used, the shaft will speed up. The gov-
High pressure header Speed controller (governor)
SC
Shaft speed sensing
Speed control valve
Load Cooling water Condensate
FIG. 8.38l Simple mechanical drive.
© 2006 by Béla Lipták
A noncondensing turbine is generally less expensive to buy and to operate than a condensing one, because the energy is extracted from the steam while it has a higher enthalpy, and hence, it has a smaller volume per unit of energy. This has the desirable effect of reducing the size of the turbine and, frequently, also increasing its efficiency. Because plants usually have a requirement for low-pressure steam for various loads, such as heating, the low-pressure steam generated by noncondensing turbines can often be used to advantage. (Figure 8.33b illustrated the control system to be used if the plant also has a requirement for high-pressure steam, and therefore total HP steam must be shared between the HP steam users and the steam turbine.) In designing any let-down steam turbine controls, it is important to evaluate the relative amounts of the exhaust steam flow from the turbine and the demand for low pressure steam in the plant and make sure that one has considered all possible combinations. Optimized Electricity Recovery For example, if the plant’s demand for low-pressure steam is variable, it is desirable to send that variable amount of steam through a let-down turbine and to recover its energy content in the form of electricity. On the other hand, one should not send more high-pressure steam to the turbine than the amount of low-pressure steam demanded by the process. These two goals can be converted into two control loops as shown in Figure 8.38m. In this control configuration, the pressure controller (PC) serves to make sure that all low-pressure steam users in the plant are always satisfied, because if the LP steam pressure
8.38 Steam Turbine Controls
High pressure steam
FT
HP steam users
FC
PC
Electricity
PT LP steam users
FIG. 8.38m This control system will follow the variable low-pressure steam demand, while sending most of the HP steam through the turbine to convert its energy content into electricity.
drops, it opens up the turbine bypass to the HP steam header. If the LP steam users cannot tolerate superheating of their steam supply, the bypass has to be provided with a desuperheater. The task of the flow controller (FC) is to make sure that whatever happens to be the LP steam demand of the plant, it is satisfied mostly by exhaust steam, from which the excess energy has already been recovered in the form of electricity. The flow controller does that by keeping the flow in the bypass at some minimum rate and increasing the HP flow to the turbine as soon as the bypass flow starts to increase. The control configuration in Figure 8.38m can only be used when the load on the turbine is not determined by the process, but is freely variable, such as in plants where the
excess steam energy is utilized for the cogeneration of electricity. While energy conservation dictates that the flow through pressure let-down line be minimized, control dynamics suggest that it should not be completely eliminated. This is because the speed of response of a let-down valve is much faster than that of a turbine. Therefore, the sensitive control of the LP steam pressure is provided by the let-down pressure controller, while the bulk of the steam passes through the turbine and is used to make electricity. Under the discussion of extraction turbine controls in a later paragraph, Figure 8.38r describes the control configuration required if the LP load of the plant can exceed the full capacity of the turbine or when the demand for LP steam can drop below the steam flow from the let-down section of the turbine. Valve Position-Based Optimization It was noted in connection with Figure 8.38m that if the low-pressure header is supplied by just throttling the high-pressure steam through a control valve, a considerable amount of energy is lost. It was shown that if the LP steam is supplied by a noncondensing turbine, much less energy is lost. In that configuration, the low-pressure header is supplied preferentially by the steam turbine. Figure 8.38n illustrates the control system for an installation where the flow controller (FC) sets the turbine’s speed controller and the objective is to maintain shaft speed relatively constant. In such a case, the amount of LP steam available from the turbine depends on its load. Therefore, if LP demand exceeds the turbine’s ability to supply it, additional steam has to be supplied through the pressure let-down valve in the turbine’s bypass. This bypass valve is controlled by the LP header pressure controller (PC).
Resets the HPS pressure controller setpoint High pressure steam (HPS) VPC 90%
Set for min. flow
Pressure let–down valve
FC
SC SP
PC Pressure controller FT
Load (electric generator) Low pressure header PT pressure transmitter Low pressure steam (LPS) Low pressure loads Condensate
FIG. 8.38n Back-pressure turbine control system for the generation of LP steam, provided with valve position-based optimizer.
© 2006 by Béla Lipták
2145
2146
Control and Optimization of Unit Operations
If the only user of high-pressure steam in the plant is the turbine shown in Figure 8.38n, then the HPS pressure controller (not shown) set point can be adjusted to keep the steam governor valve always nearly open (90%). This is done by an integral-only valve position controller (VPC) that reduces the HPS pressure controller set point whenever the governor valve is less than, say, 90% open. As the HPS supply pressure is reduced, more HP steam will be needed to meet the same electric load on the turbine, and therefore, the governor valve will open. This control strategy, which keeps the governor valve nearly open, is an optimization strategy, because the same load is being met with less pressure drop through the governor valve, and therefore it is being met at a higher efficiency. The flow controller in the turbine bypass, as was explained in connection with Figure 8.38m, serves to make sure that most of the steam is sent through the turbine. Therefore, whenever excessive amounts of steam pass through the let-down valve, the flow controller (FC) increases the speed set point and thereby increases the amount of steam passing through the turbine. Extraction Turbine Control In addition to the governor valve, in extraction turbines, a second “valve” is required. It controls the steam flow rate that is extracted from the first stage of the turbine and is sent to the second stage. The extraction rate can be controlled to keep the pressure of the LP header constant, but it can also be a function of shaft speed, or a combination of the two. If the turbine incorporates the controls as a built-in feature, the turbine is referred to as an “automatic-extraction” type. Such turbines are generally designed to deliver 100% shaft power and to provide extraction steam only if the load requirements permit. This is the most common type of extraction machine. Extraction turbines may be visualized as two-stage units from which steam can be removed at a pressure between that of the supply and that of the exhaust. When the demand for work (load) on the turbine is small, the high-pressure stage may be adequate to meet the “work load,” and consequently, a large amount of extraction steam may be available to supply the low-pressure header. As work load increases, the second stage becomes necessary to meet the demand for added work and begins to compete for the steam previously being extracted. The control system must allow for this to occur, if meeting the work load is the first priority. At least a minimal amount of steam must be maintained through the second stage to prevent overheating. This requirement may necessitate limiting extraction, but it can also require the maintaining of a specific second-stage discharge pressure. These requirements are given in the manufacturer’s operating specifications. LPS Demand Exceeding First-Stage Exhaust Figure 8.38o shows an extraction turbine in a pressure let-down application.
© 2006 by Béla Lipták
High pressure header PC SC
Inlet valve
1 st Stage
PT
2 nd Stage
Load
Extraction valve (Generally built into the turbine) Low pressure header Condensate Condensate
FIG. 8.38o The controls of an extraction turbine in a pressure let-down application, where the demand for LPS always exceeds the steam available from the interstage of the turbine.
In this example, the exhaust of the first stage is used to supply a low-pressure header, while the second stage is condensing. The speed controller (governor) is arranged by hydraulic or mechanical linkage to close (or nearly close) the extraction valve, if the turbine speed can be maintained by the governor throttling the HP steam supply valve to the first stage (the steam inlet valve). If speed cannot be maintained by the first stage alone, the extraction valve starts to open, admitting more steam to the second (condensing) stage and, consequently, starving the low-pressure header. When the available extraction steam is insufficient, similarly to the arrangement in the previous figures, a pressure-controlled bypass valve is used to maintain the pressure in the low-pressure header. LPS Demand Less Than First-Stage Exhaust So far, it has been assumed that the low-pressure header can use all the steam that is available from the turbine. If this is not the case — if the interstage steam supply is in excess of the lowpressure header’s requirements — the excess steam must be condensed or vented to protect from overpressuring the LPS header. Figure 8.38p illustrates the controls that will protect against either unnecessary condensing or wasting treated water by venting the steam. In this control configuration, if the low-pressure header does not need the steam, the HP steam supply flow to the first stage of the turbine is reduced. Because this reduces the energy available from the first stage, the extraction valve is opened to the condensing stage to supply the additional horsepower required to maintain shaft speed. The main
8.38 Steam Turbine Controls
D/A PC
−
High pressure (HPS)
Σ
Supply valve FC
PY +
2nd Condensing stage
1st Stage
Loads
SC R/A FC Automatic extraction valve
Set for minimum flow
FC
R/A − +
PY
Σ
High pressure Supply 2nd (HPS) valve Condensing stage FC 1st Stage SC R/A
FC
PC R/A PT
2147
Loads FC Nonautomatic extraction valve Low pressure (LPS) Condensate
PT Low pressure (LPS)
FIG. 8.38q The addition of a pressure-controlled let-down line increases the speed of response, while the flow controller minimizes the energy waste through that line.
Condensate
FIG. 8.38p The controls of an extraction turbine in a pressure let-down application, where the demand for LPS is always less than the steam available from the interstage of the turbine.
difference with the controls in Figure 8.38o is that the pressure controller (PC) that maintains the pressure in the LPS header does not modulate a bypass valve, because there is no bypass. Instead, it reduces the HP steam supply to the turbine if the LPS pressure rises, by sending a negative bias to the summing relay (PY). The speed controller (SC) in this configuration controls the extraction valve to maintain the shaft speed. The speed controller’s output to the HPS supply valve is sent to the positive input of a summing relay (PY). In the absence of a signal from the PC to the negative input of PY, this scheme would operate the same way as the controls in Figure 8.38o. The pressure controller on the LPS header in Figure 8.38p is direct-acting, so its output will increase when the LPS header pressure rises. This increasing control signal will be subtracted from the speed controller’s output and will cause the FC inlet valve to close slightly. This, in turn, will cause the turbine to slow down. As the shaft speed drops, the speed controller will attempt to open the inlet valve by increasing its output signal, but the change will again be resisted by the pressure controller. As speed falls off, the reverse-acting speed controller will open the extraction valve. As a consequence, the pressure in the LPS header will decrease. This will cause the pressure controller to reduce its output, which in turn will slightly open the HPS supply valve, which will increase the shaft speed, and therefore, the speed controller will close the extraction valve somewhat. Eventually, after much interaction, a new equilibrium will be achieved. Improving the Control Dynamics The dynamics of the pressure control system in Figure 8.38p is dependent on the
© 2006 by Béla Lipták
dynamics of the turbine’s extraction control system. The characteristic of the extraction valve is often not suitable for throttling control and can be improved by use of a characterizing positioner. Another method of improving the dynamic response of the control system in Figure 8.38p is to provide a pressure let-down bypass line, as shown in Figure 8.38q. Here, the pressure controller (PC) provides the sensitivity required for quick response, while the flow controller (FC) keeps the flow in the bypass line at a minimum, as it slowly opens the HPS supply valve to the turbine whenever the flow in the bypass line exceeds its set point. Flexibility by Controller Sequencing If the relative sizes of the work load on the turbine and the user’s demand for LPS are unpredictable, the previously described control systems will not work. Figure 8.38r illustrates a control strategy that utilizes controller set-point sequencing to allow optimized and stable operation under any combination of relative load sizes. If the demand for LPS exceeds the amount of exhaust steam available, the control system shown in Figure 8.38r will operate in a similar manner as did the control system in Figure 8.38o, by the pressure controller (PC) providing the additional steam through the pressure let-down bypass line. The main difference between the two control systems is the addition of the optimizing controller FC-3, which is set at a bypass flow rate slightly exceeding the set point of FC-1. Therefore, if the pressure controller (PC) opens the bypass and the let-down flow rate exceeds the set point of FC-1, the previously inactive (saturated) FC-3 becomes active and starts cutting back the LPS steam flow to the boiler feed water preheater and, thereby, reduces the plant’s demand for LPS. This is an energy-efficient response because the energy recovered from the LPS supplied to the feedwater preheater is less than the energy content of the HP steam that is needed to produce that LPS.
2148
Control and Optimization of Unit Operations
HP steam users
High pressure steam SP = A FT SP = A + 1% FC 3
FC 1 FC 2
CV−1 SP = A − 1%
PC
1st stage
2nd stage
Load
Condensate PT Feed water preheater
CV−3
CV−2
LP steam users
FIG. 8.38r This control system is both flexible and optimized: FC-1 keeps the bypass flow to a minimum, while FC-3 reduces the LPS demand if it exceeds the work load on the turbine, and FC-3 makes more steam energy available to the turbine if the LPS demand is below the work load.
If the relative loads are reversed and the LPS availability exceeds the demand for LPS, this will cause the pressure in the LPS header to rise and the pressure controller (PC) to reduce down the bypass flow. When this let-down flow drops below the set point of FC-1, the previously inactive (saturated) FC-2 becomes active and admits that part of the LP steam that is not needed in the LPS header into the second, condensing stage of the turbine. Any number of such bypass flow controllers can be used to sequentially respond to changes in the relative sizes of the work and LP steam loads. These controllers should be provided by integral action only, so that they will be saturated (and their control valves closed) until their set points are reached. Decoupling the Interaction If the turbine is a nonautomatic extraction type, and therefore one can send a control signal to the extraction valve (as was the case in Figure 8.38r), the interaction between the pressure and turbine speed controllers can be decoupled. In the control system configuration shown in Figure 8.38s, a drop in the speed of the shaft opens both the inlet and the extraction valves, and an increase in shaft speed closes both of them. In this control system, when the LPS header pressure rises, the pressure controller (PC) output rises, and therefore the PY-1 output drops and the supply valve closes. At the same time, the increase in the PC output increases the output of PY-2, which opens the extraction valve. When the LPS header pressure drops, the opposite is the response: The supply valve opens and the extraction valve closes. If the weighing of the combining algorithms (PY-1 and PY-2) is correct (if they properly consider features of the valves and the time constants of the loop components), the response to changing pressure conditions, the supply, and extraction valves will complement each other.
© 2006 by Béla Lipták
Therefore, if the control model is properly tuned and the gains of the summers PY-1 and PY-2 are properly set, there will be no interaction between speed and pressure control,
High pressure header Supply valve FC
1st Stage
2nd Stage
Loads
SC R/A FC Nonautomatic extraction valve
+ PY − 1
+ PY Σ 2 +
PC
D/A
PT Low pressure header Condensate
FIG. 8.38s One way to eliminate interaction between flow and pressure loops is to allow the pressure controller to throttle both the supply and the extraction valves.
8.38 Steam Turbine Controls
2149
damage. Thermal stress monitoring performs the calculations needed to determine thermal conditions of the turbine and safe parameters for control operation. It also provides the operator with information on rotor thermal stress (acceleration and load rates, and maximum allowable initial load pick-up). The main safety control element on a turbine is the steam supply valve. This safety valve can be a separate on/off valve, or the shut-off function can be incorporated into the controls of the steam supply valve that is used for speed control. Taking the turbine off-line is accomplished by closing this valve. Consequently, the safety control system should be so designed as to require that all interlocks be satisfied before this valve is allowed to open. As shown in Figure 8.38t, the safety interlocks usually include the safety response to lube oil failure, high bearing temperature, overspeed, and vibration. Lube oil is generally monitored by a pressure switch in the case of pressure lubrication systems or by a level switch in nonpressure systems. If the lube oil failure switch is not satisfied, the turbine shutoff valve is closed. The sudden loss of load will cause the turbine to overspeed. This can happen in mechanical drive applications, but it is a more common occurrence in electrical generator drives. Abnormal electrical conditions in a distribution system can
and the speed control will not be adversely affected by the responses to pressure disturbances, and vice versa. The speed of the response of this system can also improved by the addition of bypass let-down controls be shown in Figure 8.38q. On the other hand, if the speed of response of the turbine is sufficient to respond to pressure variations, the let-down station (shown in Figure 8.38q) can be eliminated. An intriguing aspect of this control configuration is the possibility of eliminating the need to throttle steam completely, if the turbine’s operating capacity sufficiently matches the low-pressure header’s demand for LPS. In that case the supply valve is kept fully open, as was the case in the control configuration shown in Figure 8.38n, and the pressure controls of the LPS header determine the distribution of the extracted steam between the turbine’s second stage and the LPS header. Obviously, this configuration is only viable if the size and characteristics of the turbine are properly selected.
SAFETY CONTROLS The turbine protection system protects the turbine from overspeeding, monitors all critical turbine parameters, and trips the turbine if a condition exists that could cause equipment
Coupling PD ∅
RTD
Lube oil PSL
PD
R 3Y
4X
PD
Turbine shutoff valve Steam
Turbine Active thrust
PD 6X
5Y
Power Axial supply RPM (thrust) alarm tachometer position module
Annunciator
Turbine alarm
Radial Bearing vibration high temp pressure
Radial vibration low pressure Monitoring equipment
PD
RTD
R T
1
PD
End of shaft 2
PD
FIG. 8.38t Turbine “health-monitoring” system. PD: Position detector, proximity sensor. (Adapted from Figure C-1 of American Petroleum Institute Standard 670, Noncontacting Vibration and Axial Position Monitoring System.)
© 2006 by Béla Lipták
2150
Control and Optimization of Unit Operations
cause protective devices to separate the generator from the system. In that case, a generator that may have been supplying megawatts of power and, consequently, megawatts of load to its prime mover can suddenly disappear, leaving only a small demand for energy that will serve to overcome the friction of its bearing and the windage on its moving parts. When this occurs, if the control system is not fast enough to reduce the steam supply to the turbine, it will overspeed. This condition can be detected by a variety of devices, such as centrifugal switches, electronic tachometers, or straindetecting devices installed on or near such components of the machine that are affected by the overspeed condition. Bearing temperature is an additional indication if lubrication is functioning properly. It is also a way of detecting the deterioration in bearings before complete mechanical failure. Mechanical wear accelerates as a result of improper lubrication or because of mechanical stresses that cause deformation of the bearing’s component parts. In either case, the bearing begins to dissipate abnormally large amounts of energy, which in turn results in heating. Consequently, a sudden rise in bearing temperature is generally an indication of incipient failure. It is important to quickly stop the machine when a potential bearing failure is detected. This is because some turbine designs maintain very small clearances between stationary and rotating parts. If the bearing deforms, it may mean the total destruction of the machine. Especially on larger machines, a stationary vibrationmonitoring system is usually installed. Such a system generally consists of accelerometers or proximity sensors located radially in each bearing and axially on the end of the shaft or the thrust collar. Two sensors positioned at right angles are typically used in bearings. Electronic monitoring equipment is used to measure the acceleration or displacement that occurs at each monitoring point. The monitoring equipment generally incorporates an “alarm” setting, which is intended to warn an operator of an impending problem, and a “danger” setting, which is intended to shut the machine down. In many instances, a vibration-initiated shutdown can prevent major damage in situations in which, without prompt action, equipment could be lost. Sequential Controls In power generating installations, the monitoring and sequential controls serve to automatically bring the turbine from turning gear to generator grid synchronization. These controls evaluate such parameters as bearing temperatures and vibration, water detection, and differential expansion. The control package advises the operator on current turbine status and provides recommended actions. Vibration monitoring equipment is covered extensively in Section 7.22 in Chapter 7 of the first volume of this handbook.
© 2006 by Béla Lipták
CONCLUSIONS Steam turbines are versatile energy conversion devices that, in addition to powering a variety of mechanical loads, can do an excellent job of extracting energy that might otherwise be wasted from a plant’s thermodynamic cycle. They provide opportunities for process improvements and energy savings. Therefore, their application should be carefully and insightfully considered. The controls of the high-pressure and lowpressure steam admission valves most often are implemented through the existing (furnished electrohydraulic) turbine controls. Advanced controls include the features of redundant control, on-line tuning, field-proven hardware, and remote operator displays, including custom graphics, report generation, and on-line, systemwide integration. There usually are four separate and redundant control packages, which perform operator automatic control (OAC), thermal stress monitoring, sequential control, and turbine protection functions. As has been discussed in this section, OAC control includes speed, load, and steam pressure modulating control, as well as valve testing and remote control operation (autodispatch, autosynchronizer, and process interface).
Bibliography Adamski, R. S., “Improved Reliability of Rotating Machinery,” InTech, February 1983. Anderson, P. M. and Fouad, A. A., Power System Control and Stability, New York: Wiley-IEEE Press, 2002. API Standard 611, “General-Purpose Steam Turbines for Refinery Services,” Washington, D.C.: The American Petroleum Institute, 1997. API Standard 612, “Special-Purpose Steam Turbines for Refinery Services,” Washington, D.C.: The American Petroleum Institute. API Standard 670, “Noncontacting Vibration and Axial Position Monitoring System,” 1st edition, Washington, D.C.: The American Petroleum Institute, 2000. Bachmann, R., et al., Combined-Cycle Gas and Steam Turbine Power Plants, 2nd Edition, Tulsa, OK: PennWell, 1999. Bentley, D. E., Machinery Protection Systems for Various Types of Rotating Equipment, Minden, NV: Bentley Nevada Corp., 1980. Bloch, H. P., A Practical Guide to Steam Turbine Technology, New York: McGraw-Hill, 1995. Boyce, M.P., Handbook for Cogeneration and Combined Cycle Power Plants, ASME, 2001. Champigny, R. A., “Retrofitting Turbine and Boiler Controls,” ISA/93 Technical Conference, Chicago, IL, September 19–24, 1993. Feuell, J., “Single-Stage Steam Turbine-Generator Set Replaces Pressure Reducing Station to Reduce Plant Energy Costs,” Turbomachinery International, January–February 1980. Feuell, J., “Steam Turbine Induction Generator Set,” Turbomachinery International, May–June 1982. Gires, T. C., and Birnbaum, M., “Digital Control for Large Steam Turbines,” Proceedings of the American Power Conference, Chicago, IL, April 1968. L&K International Training, Institute of Electrical & Electronics Engineering, 1999. NEMA Standard No. SM21, “Multistate Steam Turbines for Mechanical Drive Service.” NEMA Standard No. SM22, “Single-Stage Turbines for Mechanical Drive Service.”
8.38 Steam Turbine Controls
Oetkin, A., “Steam Turbine Applications of 2301 Control Systems,” Ft. Collins, CO: The Woodward Governor Co. Osborne, R. L., “Controlling Central Station Steam Turbine Generators,” Instruments and Control Systems, November 1975. Podolsky, L. B., “A Feedforward System for Digital Electrohydraulic Turbine Control,” 1973 ISA Conference, Reprint No. 73–661, 1973. Podolsky, L. B., Osborne, R. L., and Heiser, R. S., “Digital Electrohydraulic Control for Large Steam Turbines,” Proceedings of the 14th International ISA Power Instrumentation Symposium, New York, May 1971.
© 2006 by Béla Lipták
2151
Shapiro, L. J., “Condensing Turbine Can Improve Economics of Cogeneration,” Power Magazine, Vol. 126, No. 8, pp. 73–74, August 1982. Still, U., and Zorner, W., Steam Turbine Governors, Process Control, John Wiley & Sons, 1996. The Woodward Governor Co., “Analytic Representation of MechanicalHydraulic and Electrohydraulic Governors,” Bulletin 25067, Ft. Collins, CO: The Woodward Governor Co. The Woodward Governor Co., “43027 Electric Control for Steam Turbine Applications,” Ft. Collins, CO: The Woodward Governor Co.
