Tuesday, April 24, 2018

IAM Search

Final Control Elements - Control Valves - Page 3

Article Index
Final Control Elements - Control Valves
Dampers and Louvres
Valve Packing
Valve Positioners
Control Valve Sizing, continued
Control Valve Problems

Valve packing

Regardless of valve type, all stem-actuated control valves require some form of seal allowing motion of the stem from some external device (an actuator) while sealing process fluid so no leaks occur between the moving stem and the body of the valve. The general term for this sealing mechanism is packing.

This mechanical feature is not unlike the stuffing box used to seal seawater from entering a boat or ship at the point where the propeller shaft penetrates the hull:

 

 

The fundamental problem is the same for the ship as it is for the control valve: how to allow a moving shaft to pass through what needs to be an impenetrable barrier to some fluid (in the case of the ship’s hull, the fluid is seawater). The solution is to wrap the shaft in a flexible material that maintains a close fit to the shaft without binding its motion. A traditional packing material for ship propeller shafts is flax. Some form of lubrication is usually provided so this packing material does not impose excessive friction on the shaft’s motion4.

Modern marine stuffing boxes use advanced materials such as Teflon (PTFE) or graphite instead of flax, which wear longer and leak less water. In the world of control valves, the traditional packing material used to be asbestos (shaped into rings or ropes, much like flax used to be shaped for use in stuffing boxes), but is now commonly Teflon or graphite as well.

In the case of a ship’s stuffing box, a little bit of water leakage is not a problem since all ships are equipped with bilge pumps to pump out collected water over time. However, leakage is simply unacceptable in many industrial control valve applications where we must minimize fugitive emissions. A “fugitive emission” is any unwanted escape of process substance into the surrounding environment, usually from leaks around pump and valve shafts. Special “environmental” packing sets are available for control valve applications where this is a concern.

Packing in a sliding-stem valve fits in a section of the valve body called the bonnet, shown in this simplified diagram of a single-ported, stem-guided globe valve:


 


Here, the packing material takes the form of several concentric rings, stacked on the valve stem like washers on a bolt. These packing rings are forced down from above by the packing flange to apply a compressive force around the circumference of the valve stem.

Two nuts threaded onto studs maintain proper force on the packing rings. Care must be taken not to over-tighten these nuts and over-compress the packing material, or else the packing will create excessive friction on the valve stem. Not only will this friction impede precise valve stem motion, but it will also create undue wear on the stem and packing, increasing the likelihood of future packing leakage.

A closer look at the bonnet shows a multitude of components working together to form a low friction, pressure-tight seal for the moving valve stem:

 


 

In this diagram, we see two sets of packing rings separated by a metal piece called a lantern ring.

The lantern ring acts as a spacer allowing lubricant introduced through the lubrication port to enter into both packing sets from the middle of the bonnet.

Photographs taken of an actual valve packing assembly removed from the bonnet (left), and reassembled on the valve stem (right) reveal the structure of the packing and associated components.

 

 

There is no lantern ring in this packing assembly, but there is a coil spring5. This makes it a live-loaded packing as opposed to a jam packing. Jam packings (without springs) must be periodically adjusted to compensate for compressive fatigue and/or wear of the packing rings.

In packing applications requiring external lubrication, a stem packing lubricator may be connected to the lubrication port on the bonnet. This device uses a long, threaded bolt as a piston to push a quantity grease into the packing assembly:

 

 

To operate a lubricator, the hand valve on the lubricator is first secured in the closed (shut) position, then the bolt is fully unscrewed until it falls out of the lubricator body. An appropriate lubricating grease is squeezed into the bolt hole in the lubricator body, and the bolt threaded back into place until hand-tight. Using a wrench or socket to tighten the bolt a bit more (generating pressure in the grease) and opening the hand valve allows grease to enter the packing chamber. The bolt is then screwed in fully, pushing the entire quantity of grease into the packing. As a final step, the hand valve is fully shut so there is no way for process liquid to leak out past the bolt threads.

The two most common packing materials in use today are Teflon (PTFE) and graphite. Teflon is the better of the two with regard to friction and stem wear6. Teflon is also quite resistant to attack from a wide variety of chemical substances. Unfortunately, it has a limited temperature range and cannot withstand intense nuclear radiation (making it unsuitable for use near reactors in nuclear power plants). Graphite is another self-lubricating packing material, and it has a far greater temperature range than Teflon7 as well as the ability to withstand harsh nuclear radiation, but creates much more stem friction than Teflon. Graphite packing also has the unfortunate property of permitting galvanic corrosion between the stem and bonnet metals due to its electrical conductivity. Sacrificial zinc washers are sometimes added to graphic packing assemblies to help mitigate this corrosion, but this only postpones rather than prevents corrosive damage to the stem.

Hybrid packing materials, such as carbon-reinforced Teflon, also exist in an attempt to combine the best characteristics of both technologies.

A completely different approach to packing is a device called a bellows seal : an accordion-like metal tube fastened to the valve stem and to the bonnet, forming a leak-proof seal with negligible friction. The accordion ribs give the bellows seal an ability to stretch and compress with a sliding stem’s linear motion. Since the bellows is an uninterrupted metal tube, there is no place at all for leaks to develop:

 

The accordion-shaped bellows is contained and protected inside the thick metal tube visible in this photograph. One end of the bellows is welded to the valve stem, and the other end is welded to the protective tube. With the tube firmly clamped in the bonnet of the valve, a leak-free seal exists.

While a properly operating bellows seal is indeed leak-free, there is always the potential of a rupture in the bellows, which could develop into a leak of catastrophic proportions. For this reason, bellows seals are almost always followed by standard packing around the stem. In the unlikely event of a bellows rupture, this standard packing will provide a serviceable seal until the bellows can be replaced.

 

4Some packing materials, most notably Teflon and graphite, tend to be self-lubricating.

5An alternative to coil-shaped springs for live-loading of valve packing are Belleville washers. These washers are made of spring steel and have a slightly concave shape, giving them resistance to compression along the shaft axis. Belleville washers are always stacked in opposed pairs.

6Based on friction values shown on page 131 of Fisher’s Control Valve Handbook (Third Edition), Teflon packing friction is typically 5 to 10 times less than graphite packing for the same stem size!

7Graphite packing is usable in services ranging from cryogenic temperatures to 1200 degrees Fahrenheit, as opposed to Teflon which is typically rated between -40o F and 450o F.



Valve seat leakage

In many process applications, it is important that the control valve be able to completely stop fluid flow when placed in the “closed” position. Although this may seem to be a fundamental requirement of any valve, it is not necessarily so. Many control valves spend most of their operating lives in a partially-open state, rarely opening or closing fully. Additionally, some control valve designs are notorious for the inability to completely shut off (e.g. double-ported globe valves). Given the common installation of manual “block” valves upstream and downstream of a control valve, there is usually a way to secure zero flow through a pipe even if a control valve is incapable of tight shut-off. For some control valve applications, however, tight shut-off is a mandatory requirement.

For this reason we have several classifications for control valves, rating them in their ability to fully shut off. Seat leakage tolerances are given roman numeral designations, as shown in this table8:

 

  Class   Maximum allowable leakage rate Test pressure drop
I (no specification given) (no specification given)
II 0.5% of rated flow capacity, air or water   45-60 PSI or max. operating  
III 0.1% of rated flow capacity, air or water 45-60 PSI or max. operating
IV 0.01% of rated flow capacity, air or water 45-60 PSI or max. operating
V   0.0005 ml/min water per inch orifice size per PSI   Max. operating
VI Bubble test, air or nitrogen 50 PSI or max. operating

 

The “bubble test” used for Class VI seat leakage is based on the leakage rate of air or nitrogen gas past the closed valve seat as measured by counting the rate of gas bubbles escaping a bubble tube submerged under water. For a 6 inch valve, this maximum bubble rate is 27 bubbles per minute (or about 1 bubble every two seconds):

 


 

It is from this leakage test procedure that the term bubble-tight shut-off originates. Class VI shut-off is often achievable only through the use of “soft” seat materials such as Teflon rather than hard metal-to-metal contact between the valve plug and seat. Of course, this method of achieving bubble-tight shut-off comes at the price of limited operating temperature range and the inability to withstand nuclear radiation exposure.

 

8Data in this table taken from Fisher’s Control Valve Handbook.

 


Control valve actuators

The purpose of a control valve actuator is to provide the motive force to operate a valve mechanism. Both sliding-stem and rotary control valves enjoy the same selection of actuators: pneumatic, hydraulic, electric motor, and hand (manual).

 

Pneumatic actuators

Pneumatic actuators use air pressure pushing against either a flexible diaphragm or a piston to move a valve mechanism. The following photograph shows a cut-away control valve, with a pneumatic diaphragm actuator mounted above the valve body. You can see the large coil spring providing default positioning of the valve (air pressure acting against the diaphragm moves the valve against the spring) and the rubber diaphragm at the very top. Air pressure applied to the bottom side of the diaphragm lifts the sliding stem of the valve in the upward direction, against the spring’s force which tries to push the stem down:

 

 

The air pressure required to motivate a pneumatic actuator may come directly from the output of a pneumatic process controller, or from a signal transducer (or converter) translating an electrical signal into an air pressure signal. Such transducers are commonly known as I/P or “I to P” converters, since they typically translate an electric current signal (I) of 4 to 20 mA DC into an air pressure signal (P) of 3 to 15 PSI.

The following photographs show I/P transducers of different make and model. A Fisher model 846 appears in the upper-left photograph, while an older Fisher model 546 appears in the upper-right (with cover removed). A Foxboro model E69F I/P appears in the lower-left photograph, while a Moore Industries model IPT appears in the lower-right:

 

 

Despite their differing designs and appearances, they all function the same: accepting an analog DC current signal input and a clean supply air pressure of about 20 PSI, outputting a variable air pressure signal proportional to the electric current input. An interesting feature to compare between these four I/P transducers is their relative ruggedness. Every transducer shown except the Moore Industries model (lower-right) is built to withstand direct exposure to a process atmosphere, hence the heavy cast-metal housings and electrical conduit fittings. The Moore Industries unit is intended for a sheltered location, and may be plugged in to a “manifold” with several other I/P transducers to form a compact bank of transducers capable of driving air pressure signals to several valve actuators.

Some pneumatic valve actuators are equipped with hand jacks which are used to manually position the valve in the event of air pressure failure. The next photograph shows a sliding-stem control valve with pneumatic diaphragm actuator and a “handwheel” on the top:

 

 

Note the three manual valves located around the control valve: two to block flow through the control valve and one to bypass flow around the control valve in the event of control valve failure or maintenance. These manual valves happen to be of the gate design, with rising-stem actuators to clearly show their status (stem protruding = open valve ; stem hidden = closed valve). Such blockand- bypass manual valve arrangements are quite common in the process industries where control valves fulfill critical roles and some form of manual control is needed as an emergency alternative.

Note also the air pressure tubing between the valve actuator and the air supply pipe, bent into a loop. This is called a vibration loop, and it exists to minimize strain on the metal tubing from vibration that may occur.

Pneumatic actuators may take the form of pistons rather than diaphragms. Illustrations of each type are shown here for comparison:

 

 

Piston actuators generally have longer stroke lengths than diaphragm actuators, and are able to operate on much greater air pressures9. Since actuator force is a function of fluid pressure and actuator area (F = PA), this means piston actuators are able to generate more force than diaphragm actuators. The combination of greater force and greater displacement yields more work potential for piston actuators than diaphragm actuators of equivalent size, since mechanical work is the product of force and displacement (W = Fx).

Perhaps the greatest disadvantage of piston actuators as applied to control valves is friction between the piston’s pressure-sealing ring and the cylinder wall. This is not a problem for on/off control valves, but it may be a significant problem for throttling valves where precise positioning is desired. Diaphragm actuators do not exhibit the same degree of friction as piston actuators because the elastic diaphragm rolls and flexes rather than rubs against a stationary surface as is the case with piston sealing rings.

A double-piston pneumatic actuator appears in the next photograph, providing the mechanical force needed to turn an on/off butterfly valve:

 

 

In this particular actuator design, a pair of pneumatically-actuated pistons move a rack-and-pinion mechanism to convert linear piston motion into rotary shaft motion to move the butterfly trim. Note the rotary indicator (yellow in color) at the end of the rotary valve stem, showing what position the butterfly valve is in. Note also the travel switch box (black in color) housing multiple limit switches providing remote indication of valve position to the control room.

Photographs of a cut-away rack-and-pinion piston actuator (same design, just smaller) show how the pistons’ linear motion is converted into rotary motion to actuate a rotary valve:

 

 

Another pneumatic piston actuator design uses a simple crank lever instead of a rack-and-pinion gear set to convert linear piston motion into rotary motion. This next photograph shows such a piston actuator connected to a ball valve:

 

 

Hydraulic actuators

Hydraulic actuators use liquid pressure rather than gas pressure to move the valve mechanism. Nearly all hydraulic actuator designs use a piston rather than a diaphragm to convert fluid pressure into mechanical force. The high pressure rating of piston actuators lends itself well to typical hydraulic system pressures, and the lubricating nature of hydraulic oil helps to overcome the characteristic friction of piston-type actuators. Given the high pressure ratings of most hydraulic pistons, it is possible to generate tremendous actuating forces with a hydraulic actuator, even if the piston area is modest. For example, an hydraulic pressure of 2,000 PSI applied to one side of a 3 inch diameter piston will generate a linear thrust of over 14,000 pounds (7 tons)!

In addition to the ability of hydraulic actuators to easily generate extremely large forces, they also exhibit very stable positioning owing to the non-compressibility of hydraulic oil. Unlike pneumatic actuators, where the actuating fluid (air) is “elastic,” the oil inside a hydraulic actuator cylinder does not yield appreciably under stress. If the passage of oil to and from a hydraulic cylinder is blocked by small valves, the actuator will become firmly “locked” into place. This may be a decided advantage for certain valve-positioning applications, where the actuator must resist forces generated on the valve trim by process fluid pressures.

Some hydraulic actuators contain their own electrically-controlled pumps to provide the fluid power, so the valve is actually controlled by an electric signal. Other hydraulic actuators rely on a separate fluid power system (pump, reservoir, cooler, hand or solenoid valves, etc.) to provide hydraulic pressure on which to operate. Hydraulic pressure supply systems, however, tend to be more limited in physical span than pneumatic distribution systems due to the need for thick-walled tubing (to contain the high oil pressure), the need to purge the system of all gas bubbles, and the problem of maintaining a leak-free distribution network. It is usually not practical to build a hydraulic oil supply and distribution system large enough to cover the entirety of an industrial facility. Another disadvantage of hydraulic systems compared to pneumatic is lack of intrinsic power storage. Compressed air systems, by virtue of air’s compressibility (elasticity), naturally store energy in any pressurized volumes, and so provide a certain degree of “reserve” power in the event that the main compressor shut down. Hydraulic systems do not naturally exhibit this desirable trait.

A hydraulic piston actuator attached to a large shut-off valve (used for on/off control rather than throttling) appears in the next photograph. Two hydraulic cylinders may be seen above the round valve body, mounted horizontally. Like the pneumatic piston valve shown earlier, this valve actuator uses a rack-and-pinion mechanism to convert the hydraulic pistons’ linear motion into rotary motion to turn the valve trim:

 

 

A feature not evident in this photograph is a hydraulic hand pump that may be used to manually actuate the valve in the event of hydraulic system failure.

 

Self-operated valves

Although not a type of actuator itself, a form of actuation worthy of mention is where the process fluid pressure itself actuates a valve mechanism. This self-operating principle may be used in throttling applications or on/off applications, in gas or liquid services alike. The process fluid may be directly tubed to the actuating element (diaphragm or piston), or passed through a small mechanism called a pilot to modulate that pressure before reaching the valve actuator. This latter design allows the main valve’s motion to be controlled by an adjustable device (the pilot).

A very common application for pilot-operated control valves is gas pressure regulation, especially for fuel gas such as propane or natural gas used to fuel large industrial burners. This next photograph shows a Fisher gas pressure regulator used for regulating the pressure of natural gas fueling an industrial burner:

 

 

This next pilot-operated valve is used in a liquid (wastewater) service rather than gas:

 

 

A consumer-grade application for pilot-operated valves is irrigation system control, where the solenoid valves used to switch water flow on and off to sprinkler heads use pilot mechanisms rather than operate the valve mechanism directly with magnetic force. A small solenoid valve opens and closes to send water pressure to an actuating diaphragm, which then operates the larger valve mechanism to start and stop the flow of water to the sprinkler. The use of a pilot minimizes the amount of electrical power needed to actuate the solenoid.

A special case of self-operated valve is the Pressure Relief Valve (PRV) or Pressure Safety Valve (PSV). These valves are normally shut, opening only when sufficient fluid pressure develops across them to relieve that process fluid pressure and thereby protect the pipes and vessels upstream. Like the other self-operated valves, these safety valves may directly actuate using process fluid pressure or they may be triggered by a pilot mechanism sending process fluid pressure to the actuator only above certain pressures. Relief valves using pilots have the advantage of being widely adjustable, whereas non-pilot safety valves usually have limited adjustment ranges.

Relief valves typically have two pressure ratings: the pressure value required to initially open (“lift”) the valve, and the pressure value required to re-seat the valve (the blowdown pressure). A relief valve’s lift pressure will always exceed its blowdown pressure, giving the valve a hysteretic behavior.

This photograph shows a pressure relief valve on an industrial hot water system, designed to release pressure to atmosphere if necessary to prevent damage to process pipes and vessels in the system:

 

 

The vertical pipe is the atmospheric vent line, while the bottom flange of this PRV connects to the pressurized hot water line. A large spring inside the relief valve establishes the release pressure.

Another style of pressure relief valve appears in this next photograph. Manufactured by the Groth corporation, this is a combination pressure/vacuum relief valve assembly for an underground tank, designed to vent excess pressure to atmosphere or introduce air to the tank in the event of excess vacuum forming inside:

 

 

The purpose for a pressure/vacuum relief valve such as this is to relieve stresses applied to the walls of the tank due to difference of pressure inside and out. Large storage tanks are typically thin-wall for reasons of economics, and cannot withstand significant pressures or vacuums. An improperly vented storage tank may burst with only slight pressure inside, or collapse inwardly with only a slight vacuum inside. Combination pressure/vacuum relief valves such as this Groth unit reduce the chances of either failure from happening.

Of course, an alternative solution to this problem is to continuously vent the tank with an open vent pipe at the top. If the tank is always vented to atmosphere, it cannot build up either a pressure or a vacuum inside. However, continuous venting means vapors could escape from the tank if the liquid stored inside is volatile. Escaping vapors may constitute product loss and/or negative environmental impact, in which case it is prudent to vent the tank with an automatic valve such as this only when needed to prevent pressure-induced stress on the tank walls.

As previously mentioned, some pressure relief valves are pilot-controlled, their “lift” and “blowdown” pressures established by spring adjustments in the pilot mechanism rather than by adjustments made to the main valve mechanism. A photograph10 of a pilot-operated pressure relief valve used on a liquid petroleum pipeline appears here:

 

 

 

Electric actuators

Electric motors have long been used to actuate large valves, either in on/off mode or in throttling services. Advances in motor design and motor control circuitry has brought motor-operated valve (MOV) technology to the point where it regularly competes with legacy actuator technologies such as pneumatic.

An electric actuator appears in the next photograph, providing rotary actuation to a ball valve. This particular electric actuator comes with a hand crank for manual operation, in the event that the electric motor (or the power provided to it) fails:

 

 

The next photograph shows a different brand of valve actuator (Rotork) turning a large butterfly valve. Although there nothing visible in this photograph betrays the nature of this actuator’s signaling, it happens to be digital rather than analog, receiving position commands through a Profibus digital network rather than an analog 4-20 mA current signal:

 

 

 

Hand (manual) actuators

Valves may also be actuated by hand power alone. The following valves are all “manual” valves, requiring the intervention of a human operator to actuate:

 

 

Note the threaded stem of the left-hand valve in the last photograph pair. This stem rises and falls with the handle’s turning, providing visual indication of the valve’s status. Such an actuator is called a rising-stem design.

 

9The greater pressure rating of a piston actuator comes from the fact that the only “soft” component (the sealing ring) has far less surface area exposed to the high pressure than a rolling diaphragm. This results in significantly less stress on the elastic ring than there would be on an elastic diaphragm exposed to the same pressure. There really is no limit to the stroke length of a piston actuator as there is with the stroke length of a diaphragm actuator. It is possible to build a piston actuator miles long, but such a feat would be impossible for a diaphragm actuator, where the diaphragm must stretch (or roll) the entire stroke length.

10This photograph courtesy of the National Transportation Safety Board’s report of the 1999 pipeline rupture in Bellingham, Washington. Improper setting of this relief valve pilot played a role in the pipeline rupture.

 

Valve failure mode

An important design parameter of a control valve is the position it will “fail” to if it loses motive power. For electrically actuated valves, this is typically the last position the valve was in before loss of electric power. For pneumatic and hydraulic actuated valves, the option exists of having a large spring provide a known “fail-safe” position (either open or closed) in the event of fluid pressure (pneumatic air pressure or hydraulic oil pressure) loss.

The fail-safe mode of a pneumatic/spring valve is a function of both the actuator’s action and the valve body’s action. For sliding-stem valves, a direct-acting actuator pushes down on the stem with increasing pressure while a reverse-acting actuator pulls up on the stem with increasing pressure. Sliding-stem valve bodies are classified as direct-acting if they open up when the stem is lifted, and classified as reverse-acting if they shut off (close) when the stem is lifted. Thus, a sliding-stem, pneumatically actuated control valve may be made air-to-open or air-to-close simply by matching the appropriate actuator and body types. The most common combinations mix a direct-acting valve body with either a reverse- or direct-acting valve actuator, as shown in this illustration:

 

 

Valve fail mode may be shown in instrument diagrams by either an arrow pointing in the direction of failure (assuming a direct-acting valve body where stem motion toward the body closes and stem motion away from the body opens the valve trim) and/or the abbreviations “FC” (fail closed) and “FO” (fail open). Other failure modes are possible, as indicated by this set of valve symbols:

 


 

In order for a pneumatic or hydraulic valve to fail in the locked state, an external device must trap fluid pressure in the actuator’s diaphragm or piston chamber in the event of supply pressure loss.

Valves that fail in place and drift in a particular direction are usually actuated by double-acting pneumatic piston actuators. These actuators do not use a spring to provide a definite fail mode, but rather use air pressure both to open and to close the valve. In the event of an air pressure loss, the actuator will neither be able to open nor close the valve, and so it will tend to remain in position. If the valve is of the globe design with unbalanced trim, forces exerted on the valve plug will move it in one direction (causing drift).

It is important to note how the failure mode of a valve is often linked to its control action (air-to-open, air-to-close)11. That is, an air-to-open pneumatic control valve will fail closed on loss of air pressure, and visa-versa. This is an important fact because good safety engineering demands that the risk factors of the process determine proper valve failure mode rather than control system convention or habit. People may find it easier to understand the operation of an air-to-open control valve than an air-to-close valve (more signal = more process fluid flow), but this should not be a guiding principle in valve selection. Air-to-open control valves fail closed by their very nature (unless equipped with automatic “locking” devices in which case they will fail in place), which means they are appropriate for a particular process control application only if that process is safer with a failed-closed valve than with a failed-open valve. If the process is safer with a fail-open valve, then the pneumatically-actuated control valve specified for that application needs to be air-to-close.

In fact, this basic principle forms the basis – or at least it should form the basis – of decisions made for all instrument actions in critical control loops: first determine the safest mode of valve failure, then select and/or configure instrument actions in such a way that the most probable modes of signal path failure will result in the control valve consistently moving to that (safest) position. For example, consider this automated cooling system for a large power-generating engine:

 

 

Clearly, it is more hazardous to the engine for the valve to fail closed than it would be for the valve to fail open. If the valve fails closed, the engine will surely overheat from lack of cooling. If it fails open, the engine will merely run cooler than designed, the only negative consequence being decreased efficiency. With this in mind, the only sensible choice for a control valve is one that fails open (air-to-close).

However, our choices in instrument action do not end with the control valve. How should the temperature transmitter, the controller, and the I/P transducer be configured to act? In each case, the answer should be to act in such a way that the valve will default to its fail-safe position (wide open) in the event of the most likely input signal fault.

Stepping “backward” through the control system from the valve to the temperature sensor, the next instrument we encounter is the I/P transducer. Its job, of course, is to convert a 4-20 mA current signal into a corresponding pneumatic pressure that the valve actuator can use. Since we know that the valve’s failure mode is based on a loss of actuating air pressure, we want the I/P to be configured in such a way that it outputs minimum pressure in the event of an electrical fault in its 4-20 mA input signal wiring. The most common fault for a current loop is an open, where current goes to 0 milliamps. Thus, the configuration of the I/P transducer should be direct, such that a 4 to 20 mA input signal produces a 3 to 15 PSI output pressure, respectively (i.e. minimum input current yields minimum output pressure).

The next instrument in the loop is the controller. Here, we want the most likely input signal failure to result in a minimum output signal, so the valve will (once again) default to its “fail safe” position. Consequently, we should configure the controller for direct action just like we did with the I/P transducer (i.e. a decreasing PV signal from a broken wire or loose connection in the input circuit results in a decreasing output signal).

Finally, we come to the last instrument in the control loop: the temperature transmitter (TT). As with most instruments, we have the option of configuring it for direct or reverse action. Should we choose direct (i.e. hotter engine = more mA output) or reverse (hotter engine = less mA output)? Here, our choice needs to be made in such a way that the overall effect of the control system is negative feedback. In other words, we need to configure the transmitter such that a hotter engine results in increased coolant flow (a wider-open control valve). Since we know the rest of the system has been designed so a minimum signal anywhere tends to drive the valve to its fail-safe mode (wide open), we must choose a reverse-acting transmitter, so a hotter engine results in a decreased milliamp signal from the transmitter. If the transmitter has a sensor “burnout” mode switch, we should flip this switch into the low-scale burnout position, so a burned-out sensor will result in 4 mA output (low end of the 4-20 mA scale), thus driving the valve into its safest (wide-open) position.

Such a configuration – with its air-to-close control valve and a reverse-acting transmitter – may seem strange and counter-intuitive, but it is the safest design for this engine cooling system. We arrived at this “odd” configuration of instruments by first choosing the safest control valve failure mode, then choosing instrument actions in such a way that the most likely signal-path failures anywhere in the system would result in the same, consistent valve response. Of course it should go without saying that accurate documentation in the form of a loop diagram with instrument actions clearly shown is an absolutely essential piece of the whole system. If the safety of a control system depends on using any “non-standard” instrument configurations, those configurations had better be documented so those maintaining the system in the future will know what to expect!

 

11Exceptions exist for valves designed to fail in place, where a valve may be engineered to “lock” in position through the action of an external device whether the valve itself is air-to-open or air-to-close.

 


Actuator bench-set

Valve actuators provide force to move control valve trim. For precise positioning of a control valve, there must be a calibrated relationship between applied force and valve position. Most pneumatic actuators exploit Hooke’s Law to translate applied air pressure to valve stem position.

 

F = kx

 Where,

  F = Force applied to spring in newtons (metric) or pounds (British)

  k = Constant of elasticity, or “spring constant” in newtons per meter (metric) or pounds per foot (British)

  x = Displacement of spring in meters (metric) or feet (British)

 

Hooke’s Law is a linear function, which means that spring motion will be linearly related to applied force from the actuator element (piston or diaphragm). Since the working area of a piston or diaphragm is constant, the relationship between actuating fluid pressure and force will be a simple proportion (F = PA). By algebraic substitution, we may alter Hooke’s Law to include pressure and area:

 

F = kx
 
PA = kx

Solving for spring compression as a function of pressure, area, and spring constant:

 

 

When a control valve is assembled from an actuator and a valve body, the two mechanisms must be coupled together in such a way that the valve moves between its fully closed and fully open positions with an expected range of air pressures. A common standard for pneumatic control valve actuators is 3 to 15 PSI12.

There are really only two mechanical adjustments that need to be made when coupling a pneumatic diaphragm actuator to a sliding-stem valve: the stem connector and the spring adjuster. The stem connector mechanically joins the sliding stems of both actuator and valve body so they move together as one stem. This connector must be adjusted so neither the actuator nor the valve trim prevents full travel of the valve trim:

 

 

Note how the plug is fully against the seat when the valve is closed, and how the travel indicator indicates fully open at the point where the actuator diaphragm nears its fully upward travel limit. This is how things should be when the stem connector is properly adjusted.

If the stem connector is set with the actuator and valve stems spaced too far apart (i.e. the total stem length is too long), the actuator diaphragm will bind travel at the upper end and the valve plug will bind travel at the lower end. The result is a valve that cannot ever fully open:

 

 

A control valve improperly adjusted in this manner will never achieve full-flow capacity, which may have an adverse impact on control system performance.

If the stem connector is set with the actuator and valve stems too closely coupled (i.e. the total stem length is too short), the actuator diaphragm will bind travel at the lower end and the valve plug will bind travel at the upper end. The result is a valve that cannot ever fully close:

 

 

This is a very dangerous condition: a control valve that lacks the ability to fully shut off. The process in which this valve is installed may be placed in jeopardy if the valve lacks the ability to stop the flow of fluid through it!

Once the stem length has been properly set by adjusting the stem connector, the spring adjuster must be set for the proper bench set pressure. This is the pneumatic signal pressure required to lift the plug off the seat. For an air-to-open control valve with a 3 to 15 PSI signal range, the “bench set” pressure would be 3 PSI.

Bench set is a very important parameter for a control valve because it establishes the seating pressure of the plug when the valve is fully closed. Proper seating pressure is critical for tight shut-off, which carries safety implications in some process services. Consult the manufacturer’s instructions when adjusting the bench set pressure for any sliding-stem control valve. These instructions will typically guide you through both the stem connector and the spring adjuster procedures, to ensure both parameters are correctly set.

123 PSI could mean fully closed and 15 PSI fully open, or visa-versa, depending on what form of actuator is coupled to what form of valve body. A direct-acting actuator coupled to a direct-acting valve body will be open at low pressure and closed at high pressure (increasing pressure pushing the valve stem toward the body, closing off the valve trim), resulting in air-to-close action. Reversing either actuator or valve type (e.g. reverse actuator with direct valve or direct actuator with reverse valve) will result in air-to-open action.

 


Pneumatic actuator response

A limitation inherent to pneumatic valve actuators is the amount of air flow required to or from the actuator to cause rapid valve motion. This is an especially acute problem in all-pneumatic control systems, where the distance separating a control valve from the controller may be substantial:

 

 


 

The combined effect of air-flow friction in the tube, flow limitations inherent to the controller mechanism, and volume inside the valve actuator conspire to create a sluggish valve response to sudden changes in controller output signal, not unlike the response of an RC (resistor-capacitor) time-delay circuit where a step-change in voltage input results in an inverse exponential output signal.

If the pneumatic valve actuator is driven by an I/P transducer instead of directly by a pneumatic controller, the problem is lessened by the ability to locate the I/P close to the actuator, thus greatly minimizing tube friction and thus minimizing the “time constant” (τ ) of the control valve’s response:

 

 


 

Still, if the pneumatic actuator is particularly large in volume, an I/P transducer may experience trouble supplying the necessary air flow rate to rapidly actuate the control valve. Certainly the problem of time delay is reduced, but not eliminated, by the close-coupled location of the I/P transducer to the actuator.

One way to improve valve response in either type of system (full-pneumatic or I/P-driven) is to use a device known as a volume booster to source and vent compressed air for the valve actuator. A “volume booster” is a pneumatic device designed to reproduce a pneumatic pressure signal (1:1 ratio), but with far greater output flow capacity. In electrical terms, a volume booster is analogous to a voltage follower: a circuit designed to boost current to a load, without boosting or diminishing voltage. A 3 to 15 PSI pneumatic pressure signal applied to the input of a volume booster will result in an identical output signal (3 to 15 PSI), but with greatly enhanced flow capacity.

A pneumatic control system equipped with a volume booster would look something like this:

 

 

Of course, enhanced air flow to and from the actuator does not completely eliminate time delays in valve response. So long as the flow rate into or out of an actuator is finite, some time will be required to change pressure inside the actuator and thus change valve position. However, if the actuator volume cannot be reduced for practical reasons of actuating force (larger diaphragm or piston area needed for more force, also resulting in more volume for any given stroke length), then the only variable capable of reducing time lag is increased air flow rate, and a volume booster directly addresses that deficiency.

Internally, a volume booster’s construction is not unlike a manually-adjusted pressure regulator13:

 

In either mechanism, an internal diaphragm senses output pressure and acts against a restraining force (either a spring preloaded by a hand adjustment screw or an external pressure signal acting on another diaphragm) to position an air flow throttling/venting mechanism. If the output pressure is less than desired, the diaphragm moves down to open the air sourcing plug and supply additional air volume to the output. If the output pressure is greater than desired, the diaphragm moves up to shut off the sourcing plug and open up the venting port to relieve air pressure to atmosphere.

 

13The volume booster design shown here is loosely based on the Fisher model 2625 volume boosting relay.

 



Promotions

  • ...more

Disclaimer

Important: All images are copyrighted to their respective owners. All content cited is derived from their respective sources.

Contact us for information and your inquiries. IAMechatronics is open to link exchanges.

IAMechatronics Login