Continue reading »]]> Work on the WirelessHART specification began over seven years ago. Since that time the WirelessHART specification has been published by the HART Communications Foundation. Also, in 2010 WirelessHART was approved by IEC as an international standard,  IEC 62591Ed. 1.0. From the very beginning, the WirelessHART design included the features required for both monitoring and control applications. When WirelessHART field devices are used in closed loop control, the PID must be modified to utilize the slow periodic or non-periodic exception reporting of measurement values. The PID in DeltaV v11 (released last year), includes the PIDPlus option that may be selected when using a WirelessHART transmitter in closed loop control.

Information on the PIDPlus feature and the performance that it enables when using wireless transmitters for control are addressed in the following:

• PID Advances in Industrial Control, IFAC Conference on Advances in PID Control PID’12, Brescia, Italy, 28-30 March 2012
• DeltaV v11 PID Enhancements for Wireless, DeltaV Whitepaper, August, 2010
• WirelessHART Successfully Handles Control, Chemical Processing, January, 2011
• Wireless – Overcoming Challenges of PID Control& Analyzer Applications, InTech, July/August, 2010
• PIDPlus An Enhanced PID Control Algorithm for Wireless Automation, AS-74.3199 Wireless Automation, Aalto University, Finland
• Incorporating Wireless Devices into Single-Use Disposable Bioreactor Design, 2009 Dhirubhai Ambani Life Sciences Symposium
• Improving PID Control with Unreliable Communications, ISA EXPO Technical Conference, 2006.
• Similarity-based Traffic Reduction to Increase Battery Life in a Wireless Process Control Network, ISA EXPO2005, Houston, TX

Ove the next month, this blog series on Control Using Wireless Devices will explore the PIDPlus technology and the control performance that it enables using WirelessHART devices.

Continue reading »]]> In January I posted a series of four (4) blogs that addressed the implement of an on-line compressor efficiency calculation. The example for this series was a compressor for a refrigeration system that used R134a refrigerant. If you have a similar compressor application in your plant then the information provided in these blogs may act as a guide in implementing this on-line calculation. In most cases, the tools commonly available in a distributed control system are sufficient to implement this on-line calculation. However, if you do not have time to implement this calculation for the variety of compressors and/or turbines in your plant then there are other ways to evaluate turbine or compressor operation efficiency that may prove to be easier to implement and support. One novel approach is based on cloud computing. MDC Technology (acquired by Emerson in Dec 2000) pioneered this approach long before Microsoft and other software companies discovered and started to promote the value of cloud computing.

This cloud based performance monitoring capability is marketed today (by Emerson) as AMS Performance Monitor. Web technology is used to collect, model, process and present performance information about critical plant equipment as illustrated below:

This capability may be used to monitor compressors (centrifugal and reciprocating) and calculate compressor efficiency (polytropic and adiabatic) as well as other compressor KPIs. These calculations are based on the American Society of Mechanical Engineers (AMSE) Performance Test Codes (PTCs) – i.e. PTC 10 for compressors. An example of the web interface to the compressor efficiency calculation is shown below.

Such a capability can be used to guide and improve plant operations.

Continue reading »]]> The week of January 8th, 2012 I traveled to our office in Costa Rica to teach the 9025 class. This class was created last year by Emerson’s education department to provide training on the basics of process control and instrumentation. The lecture material and student workshops included in the class are based on the book Control Loop Foundation – Batch and Continuous Processes. This class is normally scheduled for 4 ½ days. However, for the two classes I taught in Costa Rica, each class was condensed into 2 days. In two days it is possible to cover all the terms, concepts and techniques that a control engineer normally uses in the process industry. The students seemed to get a lot out of doing the workshops that are included in the book and on the book web site. Also, in this class I spent some time covering the chapter in the book on the development of dynamic simulations to support control system checkout and operator training. Many of the students in the class were especially interested in this area since they work on development of process simulations within the operator training group.

There were 20 students in each class. The training class room in Costa Rica and students in the second 9025 class are shown below.

It was a real pleasure working with the students in Costa Rica.  The city of San Jose is quite large. However, in most areas of the city some of the rich vegetation native to this climate has been preserved. Below is a picture showing a view of the city from my hotel window.

Continue reading »]]> Part 1 of the centrifugal compressor efficiency series addressed the measurements and parameters that must be calculated to determine compressor efficiency. In Part 2-3 of this series we examined one way of calculating the enthalpy and entropy of the gas stream at the compressor suction and discharge and isentropic enthalpy for the refrigerant R134a. Once these parameters are known, then the calculation of the compressor efficiency is very straight forward as shown in Part 1 of this series. Below is an on-line view of the completed compressor efficiency calculation for the refrigerant R134a.

When utilizing this module for on-line calculation of compressor efficiency in a DeltaV control system, the parameters shown for suction and discharge temperature and pressure would be changed to external references to the measurement values for suction and discharge pressure and temperature.

Later this year I will contribute the module shown above and associated documentation to the Emerson Application Exchange web site. Thus, through this web site the module will be freely available for download. If you have a DeltaV system and are responsible for a refrigeration compressor in your plant then I hope you find this module of value in improving your plant operation.

Continue reading »]]> A compressor going into stall is like jumping off a cliff with a bungee cord. If the bungee cord has no losses to dampen the oscillation, we have something akin to surge. A 0.5% drop in efficiency can occur for each surge cycle. Several surge cycles can occur due to delays and lags in high temperature, thrust, and vibration shutdown systems. In some compressors the damage is so severe after multiple surge cycles that rotors and seals need to be replaced. The cost of process downtime can be significant particularly when a compressor feeds parallel trains of equipment. The restart of exothermic fluidized bed reactors in the petrochemical industry may be the most hazardous mode of operation.

A precipitous drop in flow occurs in less than 0.05 seconds at the start of full surge. The oscillation period of 1 – 2 seconds is too fast for recovery by closed loop control. An open loop backup is needed to prevent a compressor trip. The culprit is what you don’t see on a compressor map.

Compressor maps typically show only the positive flow  negative slope portion of the characteristic curve possibly including the zero slope point. To the left of the zero slope is a positive slope portion preceded by a negative flow negative slope as depicted in Figure 3-1 in the excerpt Description-of-Surge from the Momentum Press book Axial and Centrifugal Compressor Control  

The negative flow negative slope can be measured by compressor manufacturers by feeding gas backwards from the discharge of the compressor. The positive slope section may be simply a 3^rd order polynomial fit between the negative flow negative slope and positive flow zero slope point of the characteristic curves. Users typically don’t get to see the curve to the left of the zero slope point that is the source of dynamic instability.

The precipitous drop in flow occurs when a discharge valve is closing and the operating point reaches the zero slope point. The positive slope provides positive feedback so fast the flow jumps horizontally on the compressor map to the negative slope in the negative flow region. The operating point then walks down the negative flow negative slope and at the start of the positive slope jumps horizontally to a point of positive flow negative slope. The result is a surge cycle shown in Figures 3-5a and 3-5b.

How do we deal with these exciting dynamics? Besides the vibration and temperature shutdown systems we need a surge controller and an open loop backup. The open loop back up has historically been triggered by a precipitous drop in flow to step open a fail open surge valve (e.g. vent or recycle valve). The open loop back holds its output or decays its output to give time for the feedback surge controller to take over smoothly. A frequent question is how fast does the feedback control loop need to be? Companies have built a business on saying the control system must have an analog controller or a high speed microprocessor with an execution time of 0.05 seconds or faster.

If the surge control system has an air actuated surge valve, even with volume boosters the pre-stroke deadtime is 0.1 sec with 2^nd order rate limited exponential lags of 0.2 sec for large compressors. The transmitter has a minimum lag of 0.1 sec. The minimum control loop reset time is about 1 sec. A 0.1 sec PID execution time adds 10% to the integrated error. The ultimate period is about 1 sec as well. A surge oscillation of 1-2 seconds is too fast to be attenuated so the job of the surge controller is to keep the compressor from getting close to the zero slope point. Feedforward from the downstream user feed flows dropping provides a helpful preemptive action for shutdowns in a parallel train. The surge controller should have minimal overshoot but not prematurely open the surge valve. The surge valve should be fast opening and slow closing. Dynamic reset limit, high speed readback of actual position, and an analog output (AO) closing setpoint velocity limit can be helpful if properly setup and tested. An open loop back up must kick in if the surge control can’t stop the operating point from reaching the zero slope point and be fast enough to prevent a shutdown on high thrust or vibration. If configured properely the open loop backup can be done in a module with a 0.1 sec execution time. Thus today’s DCS can be used for surge detection and prevention eliminating the need for special systems.

The key to an open loop back up system preventing even the start of surge is the use of a deadtime block to create a fast  train of rates of change of flow and pressure as noted in the March 4 post “A Calculation so Simple yet so Powerful”

If the rate of change of flow is negative, the open back up takes preemptive action when the rate of change of pressure approaches zero indicating the zero slope point is imminent. In this case an incremental opening of the surge valve can be used causing less disruption to downstream users.  

As I am listening to U2′s “One” I thought I would give you a heads up that I will be moving my blog to the Control magazine website to be “One” with my Control Talk column.  

“So Long, and Thanks for All the Fish”

Continue reading »]]> Part 1 of the centrifugal compressor efficiency series addressed the measurements and parameters that are used to calculate compressor efficiency. In Part 2 of this series we examine how these measurements are used to calculate the enthalpy of the gas stream at the compressor suction and discharged for the refrigerant R134a. This calculation utilized bilinear interpolation of the thermodynamic properties published by DuPont in Technical Information T-134a – ENG. The remaining parameter that must be calculated to determine compressor efficiency is isentropic enthalpy as shown in Part 1. This parameter is depends on the gas stream entropy at the compressor suction and the compressor discharge pressure.

The following steps may be used to calculate isentropic enthalpy.

1. Calculate the entropy of the gas at the compressor suction as addressed in part 2 of the series.

2. Based on the compressor discharge pressure, find the gas temperature at the discharge pressure that represents a gas entropy that is equal to the gas at the compressor suction.

3. Using the gas temperature determined in step 2 calculate the enthalpy of gas a this temperature and the compressor discharge pressure. This is the isentropic enthalpy that is needed in the turbine efficiency calculation

The key to addressing step 2 is realize that at a given pressure the entropy decreases monotonically with gas temperature as is illustrate by the following example taken from DuPont’s Technical Information T-134a – ENG.

An example implementation of the isentropic enthalpy calculation is illustrated below:

Part 4 of this series on Centrifugal Compressor Efficiency will show how the enthalpy values calculated in Parts 2 and 3 of the series are used to implement an on-line calculation of compressor efficiency.

Continue reading »]]> We conclude with a summary on how you can avoid bursts of oscillations, get to setpoint faster, choose the right execution time and filter, coordinate the speed of loops, optimize operations without while protecting equipment, provide a consistent flow response for model predictive control, eliminate limit cycles, and improve analyzer and wireless control loops (all in less than a 1000 words).

A primary controller output can be prevented from going beyond the setpoint limits of a secondary loop driven by the output by obeying setpoint limits in cascade and remote cascade mode. Stopping the output at the setpoint limit allows a more immediate recovery when the output reverses direction. The proper use of the back calculate signal enables a bumpless transfer for mode changes and a responsive transition in override control, and prevents bursts of oscillations for slow secondary loops and slow valves. The use of PV for the back calculate combined with the dynamic reset limit or “external reset feedback”, limits the primary loop output from changing faster than the secondary loop or a slow final control element (control valve or variable frequency drive) can respond preventing a burst of oscillations for large disturbances or setpoint changes. A fast readback of  valve position or drive speed is needed by a dedicated signal or primary HART variable (PV). A secondary HART variable (SV) for valve or speed readback may not be fast enough for the dynamic reset limit.

If the setpoint tracks the PV in the manual mode, then the setpoint change in the auto mode will provide a step in the controller output from the structure “PI on error and D on PV”. If the setpoint is retained at the desired operating point in auto (no setpoint change in auto), the approach to setpoint is extremely slow unless the output is prepositioned by an ROUT mode because the change in controller output to achieve the setpoint is a ramp from integral action instead of a step from the proportional mode. For temperature loops, the integral time setting is large causing a slow rise time. If the setpoint must be retained in the PID loop, setpoint tracking of PV is not used and rise time is sacrificed. For primary loops used in traditional basic control of continuous operations, there are few setpoint changes and controller outputs are prepositioned for startup. When grade transitions, flexible manufacturing, model predictive control and real time optimization result in setpoint changes, the use of setpoint tracking PV in manual enables a smooth  transition to advanced control besides cascade control.

The total of the signal filter time and the PID module execution time should be less than 10% of the smallest integral time to prevent the integrated error from an unmeasured disturbance increasing more than 10%. The filter time should be just large enough to keep fluctuations in the controller output from exceeding the final control element (valve or variable frequency drive) deadband.

Directional velocity limits on the setpoint in an Analog Output (AO) block used in conjunction with a dynamic reset limit and the use of AO PV for the back calculate signal can provide intelligent coordination and regulation of control loop speed without the need for retuning the PID.

Directional AO setpoint velocity limits can provide a slow approach to an optimum and a fast getaway from trouble for valve position control (VPC). See “Don’t Over Look PID in APC” for the many possibilities of VPC.

Directional AO setpoint velocity limits offers quick recovery from surge conditions and a lower chance of re-entry in surge by a fast opening and slow closing of the surge valve. In the old days directional slewing rate was done on the fail open surge valve by quick exhaust valves or boosters with a higher vent rate than pressurization rate. The action of these devices was disruptive and unrepeatable posing operational, maintenance, and tuning problems.

Directional AO setpoint velocity limits offer the opportunity for loops to have the same speed of response despite different tuning and dynamics. The greatest need is commonly seen in coordination of flow loop response. For blending operations, flows are set in ratio to each other. If the setpoints of the flow loops are simultaneous driven and directional velocity limits ensure the speed of the flow response is nearly identical, the blend composition will not be upset by an unbalance in flows. The same strategy is useful for minimizing the upset from load changes and analyzer corrections of ratios for reactor feeds and for using the same model for flow response in model predictive control (MPC), particularly advantageous for minimizing duplicate MPC setup and maintenance in parallel equipment trains.

The limit cycles from deadband (e.g. valve backlash with 2 or more integrators in loops and process) and resolution or threshold sensitivity limits (e.g. valve stiction with one or more integrators in loops or process) can be killed by setting the integral deadband equal to the PV amplitude of the limit cycle. The PV amplitude depends upon operating point and usually gets larger as a valve approaches the closed position or as the product or corrosion builds up on sealing or seating surfaces and stems. The enhanced PID developed for wireless will inherently kill the oscillations if  noise does not trigger an update. A filter or threshold sensitivity setting is used to screen out noise can prevent unnecessary updates.

For wireless and analyzer loops where the time between updates is much larger than the sum of the process time constant and deadtime, the enhanced PID enables the use of a PID gain that is the inverse of the open loop gain. This PID gain provides a single correction for a setpoint change that puts the PV at the setpoint for the next update (see “Wireless – Overcoming challenges of PID control & analyzer applications” for details on the opportunity).

Continue reading »]]> As addressed in Part 1 of the centrifugal compressor efficiency series, the enthalpy of the gas stream at the compressor suction and discharge must be known to determine the compressor efficiency. Also, the isentropic enthalpy must be determined based on the gas stream entropy at the compressor suction and the compressor discharge pressure. These properties of the gas stream may be calculated based on the thermodynamic property of the gas used with the compressor. Since in this example the centrifugal compressors is part of a refrigeration systems, we assume the refrigerant is R134a. The thermodynamic properties of R134a have are published in DuPont’s Technical Information T-134a – ENG that is publically available on the web. Below is an example of the property tables supplied by DuPont for superheated R134a.

As noted in this table, enthalpy and entropy are a function of the gas pressure and temperature. Based on the information provided in the property tables for specific temperatures and pressures, the enthalpy and entropy at any pressure or temperature within the table range may be calculate. For example, in one recent application, the values from the thermodynamic table that span a temperature range of -10 degF to 200 degF and a pressure range of 10psia to 200psia were entered as array parameters in a DeltaV module. Calculation blocks were then created that utilize the thermodynamic property arrays to calculate enthalpy and entropy using bilinear interpolation based on the pressure and temperature provided as block inputs as illustrated below.

The calculation of isentropic enthalpy used in the efficiency calculation will be addressed in part 3 of this series.

Continue reading »]]> To support the check list in part 8 to get the most out of your PID, we offer some of the considerations in the use of PID features. The number of PID options and parameters can seem overwhelming at first. Hopefully the knowledge gained here will foster an appreciation of the flexibility and capability leading to a positive attitude toward the possibilities.

Most PID controllers use the ISA “Standard” form. Analog controllers used the “Series” form where the derivative calculation was done on the rate of change of the process variable (PV) in series with proportional and integral calculations. This form was principally the result of analog circuitry limitations. An advantage of the “Series” form is the inherent prevention of the effective rate time from exceeding the effective reset time preventing an instability from excessive rate action. The inherent protection is the result of an interaction factor that reduces the effective rate time and controller gain and increases the controller reset time as shown in Equations 3-6 thru 3-9 in the chapter Basic-Control from the ISA book Advanced Temperature Measurement and Control. The interaction factor becomes 1.0 if the rate time is zero. These equations and interaction factors can be used for the conversion of an analog controller’s “Series” PID tuning to a DCS “Standard” PID tuning. This chapter also provides figures and a discussion of the various forms and structures commonly offered in a modern DCS.

The PID structure mode commonly used is “PI on error and D on PV.” This structure will provide a step change in the PID output from the proportional mode for a step change in the PID setpoint. Operator or sequence initiated changes in operating point are step changes in the setpoint. This step change in the output from proportional action helps get the PV to setpoint faster, which is important for reducing startup, transition, and batch times. However, this initial kick in the output is more likely to cause overshoot of the setpoint. The selection of the structure “I on error and P and D on PV” can eliminate this overshoot but the time to reach setpoint (rise time) may be painfully slow.

There is a unified approach where tuning for maximum disturbance rejection and the structure “PI on error and D on PV” can be used to minimize rise time with no overshoot. The approach is to add a lead-lag to the setpoint change. The lag time is set equal to the reset time and the lead time is set less than or equal to ¼ of the lag time. The lead time is reduced to provide a slower approach to setpoint. Recent tests show a lead 1/10 of lag most useful. For the latest developments see the InTech Feb 2012 article ”PID tuning rules”

We need to be careful about always demanding simple solutions and just clicking on buttons. While this may seem to offer a reduced configuration effort, what it really leads to is a discredit and subsequent loss of expertise and employment of engineers and technicians responsible for the configuration and an eventual loss of automation system performance. Parameters and features are hidden from the customer to make the product seem more attractive shielding the user from complexity.

There are exceptions to every rule due to the nonlinearity and diversity in the process industries. There is nothing more frustrating to a user than a feature that cannot be enabled or disabled or a parameter that cannot be set. The long term effects are often not even recognized because  no one is left who understands what could have been and should have been creating a positive feedback of cost cutting simplistic views. I encourage users to ask for more rather than less lists of adjustable parameters and selectable options. A multi-view approach may be best. There could be an executive “overview” of the key parameters, a “flexible” view for the normal user, and an “advanced” view to expand the horizons and break paradigms. The “flexible” view should help develop field expertise. The “advanced” view should encourage creativity. These views increase our profession’s value and visibility so we don’t reach a “Point of No Return” as noted in the site entry “How to Succeed – Part 2“.

This does not mean every application is a special case. We benefit enormously from conceptual thinking and seeing the commonality (see site entry “Top Ten Limitations – Conceptual Thinking“). My goal, possibly as a result of my education as a physicist, is a unified theory or approach that cuts to essence of the solution. The virtual plant enhancement of field experience with an advanced PID offers a conceptual understanding with far reaching benefits.

Next week we continue our discussion of PID features.

Continue reading »]]> Centrifugal compressors play an important role within the process industry. In some cases compressor capacity may determine the throughput of a plant. If a compressor is used for refrigeration then any variation in compressor operation may directly impact the temperature control in key unit operations such as flash vessels and exothermic reactors. Also, the compressor drive is often major consumer of electric or steam energy. Thus, an on-line calculation of compressor efficiency can be useful in evaluating the impact of operating conditions on compressor performance and cost of operation. As illustrated below, the compressor suction and discharge pressure and temperature measurements are required to calculate dynamic compression efficiency.

The dynamic compression efficiency is calculated based these pressure and temperature measurements as shown below:

It is estimated that there are over 80,000 centrifugal chillers in operation in North America. Thus, in the upcoming series of blogs on Centrifugal Compressor Efficiency will focus on the dynamic compression efficiency of a centrifugal compressor used in a plant for refrigeration. This efficiency calculation may be implement using common tools in a DCS and put on-line within the control system. By making on-line compressor efficiency available to the plant operator as a continuously calculated value, the operator can better assess the impact of operation changes on the compressor efficiency and use this knowledge to improve plant operations.

Continue reading »]]> The missing piece for success is the center piece of the loop, the PID controller. There is an incredible offering of PID features and options. To help utilize the full potential of the PID, here is a check list as a guide. While the full aspects of the PID capability are book worthy, the following overview can get you started on the right path.

If you don’t get the valve action and control action right, nothing else matters. The controller output will ramp off to an output limit. The valve action (inc-open and inc-close) can be set in many different places, such as the PID block, analog output (AO) block, splitter block, signal characterizer block, current to pneumatic (I/P) transducer, or the positioner. Make sure the valve signal is not reversed in more than one location for an inc-close (fail open) valve. Once the valve action is set properly, the control action is set to be the opposite of the process action. The control action is reverse and direct if an increase in the PID output causes the PID process variable (PV) to increase or decrease, respectively. Verify with process engineer the valve action and process action and resulting control action required. Deferring considerations, here is the checklist without delay (pun intended since we are not going to consider here the use of a delay in the external reset feedback for deadtime compensation).

 Measurement scale covers entire operating range including abnormal conditions
 Valve action (inc-open for fail close and inc-close for fail open if action not reversed in another location)
 Control action (direct for reverse process and reverse for direct process if valve action set properly)
 Form (ISA standard)
 Obey setpoint limits in cascade and remote cascade mode
 Back calculate for bumpless transfer
 PV for back calculate in secondary loop
 Structure (PI action on error, D action on PV for most loops)
 Setpoint track PV in manual unless setpoint must be inherently saved in PID
 Setpoint limits to match process, equipment, and valve constraints
 Output limits to match process, equipment, and valve constraints
 Anti-reset windup (ARW) limits to match output limits
 Execution time less than 10% of minimum reset time
 Signal filter less than 10% of minimum reset time
 Tuned with auto tuner or adaptive tuner
 Rate time less than ½ deadtime (typically zero except for temperature loops)
 Dynamic reset limit enabled for cascade, AO velocity limits, and slow valve
 AO setpoint directional velocity limits set for blending, valve position control, and surge control
 Integral deadband > limit cycle PV amplitude from deadband and resolution or enhanced PID for wireless

The setting of all options and parameters must be verified as applicable. Simulations representative of the dynamic behavior of the process and the field automation system along with the actual configuration to form a virtual plant is advisable for testing and confirmation plus training and opening the door to process control improvement (see Exceptional Opportunities in Process Control – Virtual Plants).

In Parts 9 and 10 we will explore many of the considerations for the PID checklist items.

Continue reading »]]> In June, 2100 the International Electrotechnical Commission (IEC)  distributed a new project proposal 65E/199/NP, Devices and integration in enterprise systems; Field Device Integration (FDI) – Part 1 to 7. Voting on this proposal was completed in October, 2011 and the new project was approved. Within IEC the SC65E WG7 committee will address converting the draft FDI specifications into an international standard, IEC62769. A liaison D has been established with the FDI Cooperation, LLC – a company recently established by industry to develop a single common solution for Field Device Integration.

The first meeting of the WG7 committee was sponsored by the US and met December 12-14th, 2011 in Houston, TX. As the US Expert to this committee, I coordinate the arrangements for this meeting in Houston. The WG7 committee members are from a number of countries. Representatives from the US, Germany, UK and Japan attended this first meeting – shown below.

Work on the FDI standard will be addressed by the WG7 committee in conjunction with the liaison D to the FDI Cooperation, LLC. However, anyone is welcome to sit in on a WG7 committee meeting. The next committee meeting is scheduled for mid-2012 and will be hosted by Germany.

Continue reading »]]> The number of registered participants at Emerson Exchange, 2011 exceeded 2,900 people. In technical sessions, over 300 papers were presented on a wide variety of topics. At Emerson Exchange the DeltaV future architecture team often presents papers on field trail results associated with prototypes of new technology and applications that we have developed. For example, at the conference this year we presented some preliminary field trial results for continuous data analytics. Also, John Caldwell provided information on the new MPCPlus capability that the future architecture team has worked on with product engineering – targeted for DeltaV v12.

A technical exhibit area was open at Emerson Exchange, 2011 from Mon-Wednesday, 4-8pm, after the technical sessions are complete for the day. The latest products from Emerson Process Management are on display and thus customers have an opportunity to see and discuss these products. This year, the continuous data analytics prototype and an early version of MPCPlus were shown in the DeltaV advanced control booth in the exhibit area, as shown below.

Many customers find the information exchange at Emerson Exchange to be of value and it the reason they attend each year. If you plan on attending, then Emerson Exchange, 2012 will be held Oct 8-12, 2012 at the Hilton Anaheim Hotel in Anaheim, California. Also, Emerson Exchange, 2012 will be held May 29-31, 2012 at the Hotel Maritim in Düsseldorf, Germany.

Continue reading »]]> Given a good measurement and final control element selection, location, and installation, we move on to designing the control scheme and achieving the best loop performance. The control system should minimize process interactions and optimize process quality, efficiency, and production rate. The control loops should minimize disturbances and effectively achieve new operating points. Here is a concise overview of how to succeed.

The control engineer should confer up front with the process and mechanical engineers to ensure their design minimizes process delays, disturbances, noise, and nonlinearity as discussed in the blogs on Mixing,  UA,  Pumps and Piping,  and Equipment. 

The control system first must manage inventories of gases, liquids, and solids. Equipment pressure loops control gas inventories. Pipe pressure loops control gas, liquid, and melt inventory. Equipment level loops control liquid and solids inventory. Properly designed inventory loops can inherently provide endpoint control and hence product composition control as noted in the blog Batch vs Continuous Control and Optimization – Part 3.  The inventory controlled is a mass for purposes of material balance control but there are often constraints in terms of volume (e.g. high level). When starting up a process model or the plant, the inventory loops are some of the first loops to be commissioned to prevent levels or pressures from triggering Safety Instrumentation System (SIS) actions, opening relief devices, or sucking in equipment.

The control system must have a production rate controller to meet the dynamic demands of a flexible and efficient plant where inventories are minimized, fluctuating market demands are met real time, and advantage is taken of varying energy source costs as noted in the blog Flexible Manufacturing. For continuous operations, the production rate could be simply set by a flow loop that is the leader for the other flow loops. For batch and fed-batch operations, the production rate would be a batch cycle time controller that manipulates the scheduling of resources and feed rates. Valve position controllers can be quickly added to push production rates and minimize utility and raw material costs as discussed in the November Control article “Don’t Over Look PID in APC”.

The control system must be able to optimize product quality, minimizing the margin between supply and demand. Reducing recycle and waste particularly prevalent in startup and transitions is important in terms of environmental impact and waste treatment cost but also process efficiency and capacity. Temperature is often used as an inference of composition. The use of more online and at-line analyzers would enable better batch and continuous process quality control as noted in my December Control Talk column Process Analyzers. Analyze this!

The key to achieving these objectives is to choose the best pairing of manipulated and controlled variables per relative gain analysis, using cascade control to isolate nonlinearities (valve and process) and disturbances by a fast secondary loop, and using feedforward control for coordinating loops and rejecting load disturbances. Secondary flow loops can compensate for pressure upsets and installed valve characteristic nonlinearities. Secondary cooling and heating temperature loops can make the vessel temperature loop nearly linear. Secondary flow loops and plant wide flow feedforward control can enable a plant to reach and maintain optimum operating points as discussed in the InTech article ”Feedforward control enables flexible, sustainable manufacturing.“ Some key considerations are discussed in the answer to an ISA Mentor program question posted on ISA Interchange website How do you know when feed-forward control is needed?

So we have great loops. Are we done? Tune in next week.

Continue reading »]]> As addressed in my December 12th blog posting, James Beall and I presented Part 1 of this presentation on Friday morning at Emerson Exchange, 2011 in a “meet the expert” workshop. Part 2 was not presentation at Emerson Exchange because of time limitation. In this continuation of the presentation we address when it is possible to justify the cost associated with the installation and commissioning of multi-loop techniques such as feedforward control, ratio and override control. The steps required to commission multi-loop control strategies are discussed along with common mistakes to avoid. Input is provided on how to recognize when advanced control techniques such as MPC are needed to achieve the desired control performance. Many of the ideas discussed in this session are addressed in Control Loop Foundation – Batch and Continuous Processes.

If you are interested in learning more about these topics, then the viewer below may be used to access the workshop presentation (slides only).

Also, the viewer controls may be used to download the power point file used in this presentation.

Continue reading »]]> The increased emphasis on minimizing leakage, energy use, and low cost bid has lead to increased process variability. The valve with the lowest leakage, pressure drop, and price tag often has deadband, rangeability, resolution, and sensitivity deficiencies leading to poor control creating continual oscillations even when there are no process upsets. The sustained equal amplitude oscillations can not be eliminated by tuning. Just putting the loop in auto is enough to trigger the oscillations.         

Control valve specification sheets have 50 or more entries but typically do not have a field for deadband, rangeability, resolution, and sensitivity. In fact there is no requirement that the control valve actually respond to a change in signal. The supplier is not held accountable for whether the valve internal closure member (plug, disc, or ball) actually moves and most importantly the flow really changes for a change in signal. No wonder the wrong valve wins the low cost bid. A rotary piping valve with an actuator originally designed for on-off service can generally meet all of the specification requirements. Slapping on a smart digital positioner just intensifies the deception giving a false sense of confidence because the positioner feedback is actuator shaft and not closure member position. The actuator shaft may move even though the closure member does not move due to deadband in linkages and connections, packing fraction, shaft windup, sealing and seating friction. Even if the closure member does move, the installed flow characteristic may be so flat the change in flow is trivial.

Resolution (stick-slip) should be better than ¼ the allowable process variable error divided by the open loop gain over the whole throttle range (min to max flow). The open loop gain is the product of the valve gain, process gain, and measurement gain. The valve gain is the slope of the installed characteristic taking into account any split ranging (change in valve flow per % change in controller output). The process gain is the change in the process variable in engineering units for a change in control valve flow. The measurement gain is 100% divided by the measurement span. The deadband from backlash should not be more than twice the resolution. The threshold sensitivity the actuator and positioner should be better than half the valve resolution. The sensitivity (slope of the installed characteristic) should be greater than the resolution in % divided by the product of the process gain and measurement gain. The sensitivity should not change more than by a factor of 4 unless the nonlinearity is compensated for by signal characterization, gain scheduling, or adaptive tuning. Figures 7-48a-c on pages 411-414 of  the “Essential’ book, Essentials of Modern Measurements and Final Elements in the Process Industry show the acceptable throttle ranges to keep the slope changes within the 4:1 range. There are other things more associated with valve selection and sizing not apparent from the control valve specification sheet. I recommend you use software not only to size the valve but to compute the installed characteristic using details of the mechanical and piping design.

Statements that you can allocate 5% of the system drop to save energy assumes you don’t need any rangeability. Statements on rangeability do not generally take into account the installed characteristic, deadband, and resolution. Equations 7-19a-d on page 418 of the “Essential” book offer a more realistic view of rangeability. Valve drops more than 25% of the system drop are advisable.

Finally, here is my check list with suggested values assuming one does not have the time or info to do detailed calculations of deadband, rangeability, resolution, sensitivity, or nonlinearity requirements:

 Use smart software to size the valve
 Select location and valve type to eliminate or reduce damage from flashing
 Select location and valve type to eliminate or reduce damage from erosion
 Include swage effect from piping reducer
 Use smart software to compute and plot installed valve characteristic
 Size actuator to deliver twice the max torque or thrust required
 Specify actuator threshold sensitivity better than 0.1%
 Specify smart positioner threshold sensitivity better than 0.1%
 Tune smart positioner for application (otherwise you have a dumb positioner)
 Specify deadband less than 0.4% over the entire throttle range
 Specify resolution better than 0.2% over the entire throttle range

 Use step sizes of 0.1% and flow measurement or travel gage to test response

 Ensure valve gain > 0.5% max flow per % signal over the entire throttle range
 Ensure valve gain < 2.0% max flow per % signal over the entire throttle range

Note that for valve gain also known as flow sensitivity, the max flow is valve capacity and the split range amplification effect must be included. For 50% split range point of 2 control valves the amplification is a factor of 2. For small and large valves, the effect of size on valve gain can be mitigated by an intelligent selection of a split range point. For a large valve with 4 times the capacity of a small valve, the split range would be 0-20% for the small valve and 20-100% for the big valve. For valves on different process streams, the process gain needs to be included in the calculation of the intelligent split range point. Since operations are accustomed to a 50% split range, graphics & training are needed.

Threshold sensitivity and resolution are the smallest input change that will cause the automation device to respond. For resolution the response is a step the size of the resolution limit (stair-case or quantized response). For threshold sensitivity, once the response occurs the output change matches the input change. The response of actuators and positioners is commonly characterized by a threshold sensitivity whereas the response of a valve with friction (stiction) is characterized by a resolution assuming the slip equals the stick. Threshold sensitivity and resolution will cause a limit cycle for a controller in automatic if there is integrating action in the process or in the controller via the integral mode. Deadband will cause a limit cycle if there are more than two occurrences of integrating action (e.g. integrating process such as level and integrating action in a level controller or cascade control with integrating action in both the secondary and primary PID).

If you get too much flak, tell them this all came from the author formerly known as Greg.

Continue reading »]]> At Emerson Exchange, 2011, I had the pleasure of working with James Beall, Principal Control Consultant, Emerson Process Management to host a workshop that addresses how to get the maximum return from your control investment.

The design and commissioning of the controls associated with a continuous or batch process directly impact plant operating efficiency and production quality and throughput. In this session James and I review techniques that may be used to identify control opportunities to reduce production costs, minimize variations in product quality and to maximize production within the limits set by market demand. Starting with an assessment of control loop utilization and automatic control performance, a step by step process is outlined that may be used to identifying and addressing areas where it is possible to justified the time and material costs required to improve control performance. In particular, information is provided on how to quickly tune single loop control and to recognize when variations in control loop performance are not associated with loop tuning. An overview will be provided of tools and techniques that may be used to achieve best control performance over a wide variety of operating conditions.

If you are interested in learning more about these topics, then the viewer below may be used to access the workshop presentation given on Wednesday, October 25. This presentation is around 1 hour and 30 minutes in length.

Also, the viewer controls may be used to download the power point file used in this presentation.

Continue reading »]]> We continue this series with some helpful hints for control system design and implementation. This week and next week we look at control valve selection. There are a lot of misconceptions from sales pitches that lack an understanding of the need for a valve to have minimum backlash and maximum resolution and sensitivity. Most of this stems (pun intended) from using step sizes that are way too large. Today, the smallest step change commonly cited is 0.5%. Maybe I should be happy because the step size was 25% until we had smart positioners that could tell us how much the valve shaft actually moved. What I want are 0.1% steps over the entire throttle range to sort fact from fiction.

Common Misconceptions:
• Rotary valves provide tighter control than sliding stem valves
• Stated rangeability takes into account pressure drop, backlash, and stiction
• “High Performance Valves” (tight shutoff valves) provide high performance
• Piping valves and on-off actuators can be used for control valves
• Piston actuators provide tighter control than diaphragm actuators
• Step tests at 50% open tell the whole story

Rotary valves tend not to have as large a throttle range where the gain (sensitivity) of the installed characteristic is acceptable. Rotary valves that are “High Performance Valves” and piping valves have a markedly reduced throttle range. If you consider the effect of increased backlash and stick-slip of these valves especially near the seat, the actuator shaft used for positioner feedback may not be representative of actual internal flow element (disc, ball, or slotted plug) position due to shaft windup, and a poor inherent flow characteristic, you understand these valves are not really control valves. Similarly, the scotch yoke, rack and pinion, and link arm on-off actuators have excessive backlash or insufficient resolution for throttling service. A diaphragm actuator has the best threshold sensitivity. Higher pressure diaphragm actuators have been developed extending their thrust and torque capability. The next best actuator is a double acting piston. All should have smart positioners with 2-stage or high gain pneumatic relays. High volume spool positioners used on dampers and piping valve posing as control valves have an extremely poor threshold sensitivity requiring step changes of 2% to see a response.

To learn more about what really is important about control valves and variable speed drives read Chapter 7 on Final Element Fundamentals in the ISA book Essentials of Modern Measurements and Final Elements in the Process Industries,  and the articles “Improve Control Loop Performance”, Chemical Processing, Oct, 2007 and “Key Design Components of Final Control Elements”, InTech, March-April, 2011.

Next week I give my checklist for valve selection.

Continue reading »]]> At Emerson Exchange, 2011, I had the pleasure of working with Eric Chen, Research Associate, Pickle Research Center, University of Texas at Austin SRP and Willy Wojsznis, Senior Technologist, Emerson Process Management, to host a workshop that addresses commissioning of a skid process that is characterized by a high degree of process interaction.

The size of the process equipment used in a pilot plant often dictates little storage for buffering of interaction between process units. In the workshop we discuss a skid mounted high temperature CO2 recovery process with a high degree of process interaction. An effective tuning approach provided high performance control. Also, information is provided on a dynamic process simulation that in the future will be used to exploring various control strategies.

If you are interested in learning more about these topics, then the viewer below may be used to access the workshop presentation given on Wednesday, October 25. This presentation is around 40 minutes in length.

Also, the viewer controls may be used to download the power point file used in this presentation.

Continue reading »]]> We continue this series with some helpful hints for control system design and implementation. This week we look at  measurement location and selection. Since a control system deals with change, the prevalent theme is how to improve the detection and correction for change. In the coming weeks we will take a look at control valves and how to get the most out of your PID controller.

A control system’s job is to correct for undesired changes in the process rejecting disturbances and to make desired changes in the process achieving new setpoints. You cannot control something you cannot measure. Furthermore the control system must not be fooled by extraneous changes. Here we look at sensor location and selection for best control system performance.

(1) SENSOR LOCATION OBJECTIVES (in approximate order of importance):

a. Maximize the detection of changes in the process from disturbances and setpoint changes. For composition, pH, and temperature choose the location that shows the largest change in both directions for a positive and negative change in the ratio of the manipulated flow to the feed flow realizing there are cross sectional and longitudinal temperature and concentration profiles in pipes and equipment. Areas behind baffles or near the surface or bottom of an agitated vessel or at the outlet of inline equipment may not be as well mixed. Temperature and pH sensor and analyzer sample tip should be near the center of pipe and extend well past equipment walls. Packed and fluidized bed equipment may have uneven composition and temperature distribution from channeling of flow. A series of temperature sensors across a fluidized bed at several longitudinal distances is often necessary with averaging and signal selection to get a representative measurement and prevent hot spots. The insertion length of the thermowell should be more than 5 times the diameter of the thermowell to minimize thermal conduction errors from heat conduction along the thermowell wall between the tip and process connection. Calculations should be run with program supplied by manufacturer on the allowable maximum length in terms of preventing vibration failure from wake frequencies. Resistance temperature detectors (RTDs) are more prone to vibration failure than thermocouples (TCs). Programs today may only be looking at thermowell failure. The tip of a pH electrode must be pointed down at a 30 to 60 degree angle to prevent the internal bubble in the glass electrode from lodging in the tip or at the internal electrode.

b. Minimize noise over the whole operating range reducing extraneous changes. The real definition of measurement rangeability must take into consideration the increase of noise at extremes of the range. Noise at low flow is the principle limitation to the rangeability of a differential head meter. Sufficient straight runs upstream and downstream have a critical effect. Liquid purging can cause transients from changes in the process pressure and purge flow. A location with good mixing and a single phase will minimize fluctuations in temperature and concentration and the disruption of bubbles or solids in liquids and liquid droplets in gasses hitting temperature or pH sensors or getting into sample lines for analyzers or into impulse lines for pressure and level measurements. Pressure probes in high velocity gas streams and furnaces must be designed to minimize momentum and vacuum effects. Sensors and sample probes tips should be not be on pump suctions and should be downstream of strainers and at least 25 pipe diameters downstream of the outlet of a static mixer or heat exchanger. The distance between a desuperheater and temperature sensor should provide at least 0.2 seconds of residence time, which doesn’t sound like much until you release this corresponds to 20 feet for a steam velocity of 100 feet per second (see Straight Talk and Secret Installation Effects). Solids are more problematic in that they may not dissolve in a reasonable residence time. The use of calcium hydroxide (lime) or magnesium hydroxide as a reagent may seem inexpensive until you consider the cost of poor control and solids going downstream. A dehydrated pH electrode or a pure water sample can cause a noisy measurement due to high glass and water resistance, respectively. A non aqueous solvent can cause dehydration and excessive sample resistance. The spikes from ground potentials and electromagnetic interference (EMI) can be eliminated by wireless transmitters. RTDs are less susceptible to EMI than TCs due to a higher level signal.

c. Minimize sensor deadtime and lag by reducing transportation delays and increasing velocities. The transportation delay in a pipe or sample line is the volume divided by the flow rate or the distance divided by the velocity. The lag time of temperature and pH sensors decreases with velocity by an increase in the heat transfer and mass transfer coefficient. Fouling also decreases with velocity. A thin film can dramatically increase the lag time of a sensor especially a pH glass electrode. A liquid velocity of 5 to 7 fps has been shown to greatly reduce fouling of probes. Velocities less than 1 fps significantly increase the lag time of sensors. The velocity in even highly agitated vessels is less than 1 fps unless the probe or thermowell is near the impeller tip. The air gap between the sensor and the interior tip of the thermowell must be minimized. The tip of the sensor must touch the bottom of the thermowell and annular clearance between the TC or RTD element and thermowell wall should be less than 0.02 inches allowing for temperature expansion. An aged or dehydrated or extremely cold pH electrode or a highly acidic sample can cause a large lag time in the pH response.

(2) SENSOR SELECTION OBJECTIVES (in approximate order of importance):

a. Maximize threshold sensitivity, resolution, and repeatability over the whole operating range reducing undetected and extraneous changes . The sensitivity of RTDs is more than an order of magnitude better than TCs. The sensitivity of Coriolis meters are more than an order of magnitude greater than vortex meters. Differential head meters may have good repeatability but suffer from noise plus uncertainty from pipe inside diameter and roughness and orifice edge wear.

b. Minimize drift eliminating loss of process knowledge, running at the wrong operating point, and the need for recalibration. Drift results in an offset of the measured value from the true value. An offset can be automatically corrected by upper level loop in cascade or model predictive control. Thus loops with a cascade or remote cascade setpoint are less affected by drift. However, knowledge of the process is degraded. For example, while the offset in a flow measurement is corrected by a setpoint change in a cascade loop, the error messes up material balances (process flows), energy balances (utility flows), and online process metrics for process analysis. Flow ratio control must be corrected by a composition loop for flow measurement drift. For custody flow meters, an offset is unacceptable. Smart transmitters and advances in sensor design have in many cases reduced drift and the effect of extraneous process and ambient conditions on installed accuracy by an order of magnitude. Drift in analytical, temperature, or pH is particularly troublesome because these are upper level loops often closely related to product quality. Operations may have adjusted setpoints to compensate for offsets in upper level loops but such adjustments are ad hoc and undone by the replacement of a sensor or transmitter. When there is an operational problem, the first question is often what maintenance was done. The drift of TCs is unpredictable and can be one to two orders of magnitude larger than the drift of RTDs. The drift of new pH electrode designs from sterilization and high temperature exposure has been greatly reduced. Solid state pH reference electrodes tend to drift for hours to days after installation due slow equilibration of the reference and high reference junction potential.

c. Minimize maintenance by eliminating drift by the use of the aforementioned advances in smart transmitters and sensors and by eliminating impulse (sensing) lines, sample lines, wires, and terminations. In-line flowmeters, close coupled differential pressure and pressure transmitters, in-situ probes, retractable insertion pH electrodes, and wireless transmitters greatly reduce the time spent analyzing real or perceived problems. Once an instrument requires maintenance, the device is a suspect whenever there is a problem. Platforms should provide access to sensors and transmitters. Analyzer shelters should be used for sophisticated at-line analyzers. For maximum onstream time and reliability use middle signal selection of 3 measurements that is capable of inherently riding out a single failure of any type and eliminating unnecessary maintenance by recognition of relative performance. The use of middle signal selection is particularly important for pH.

d. Minimize nonlinearity that cannot be corrected by a smart transmitter. RTDs can be consistently linearized by the use of Callendar-Van Dusen equation eliminating the error when sensors are changed. The interchangeability error for TCs is much greater than RTDs due to greater nonlinearity and unpredictability.