The Simulator can be executed in the Standard or Advanced modes. They are the same except the Standard mode removes some features that would not be frequently or typically used except by advanced users (e.g. Hardware Resolution).

The main purpose of the Simulator is to test the suitability and performance of the control with existing and proposed tuning parameters. Parameters that specify the behaviour of the process can also be adjusted to determine how the stability and performance of the control is affected by changes in product grade, pressure, etc. Noise in the PV, Setpoint or Disturbance can be added to simulate real-world conditions to ensure the control does not unnecessarily react to these. Valve performance and deterioration (stiction and backlash) can also be simulated to determine how the control performs in non-ideal present or future conditions. The possible improvements resulting from the implementation of Feed-Forward control can be tested before implementation. They can also be quantified by comparing the “Integrated (or Average) Absolute Error” with the various possible tuning settings. Simulator even distinguishes between the real-world process value VS the valued measured digitally in the control system (normally a DCS (Distributed Control System)). The goal is to investigate and learn in the Simulator rather than in your process to reduce risk and costs. Simulator can also be used as an excellent training tool.

This is a Simulator plot with the Process tab shown (one of 7 tabs where you can enter various settings for experimentation):
Here, you can enter the main variables that define the process and the controller (range, dead time, time constant, process gain, integrator process gain for levels etc.) with or without noise. The process variables are normally determined by performing bump tests or are estimated. You then set any other values in the remaining 6 tabs and click on Calculate. The results are calculated and displayed. Nine variables are calculated and graphed. Of course, the PV (process value or measurement), setpoint and controller output are the main ones and any variable can be hidden to simplify the view:

• Setpoint - Could be constant or variable such as when simulating a slave loop and could even be noisy.
• PV in Analog Form - This is the process real value and is the primary result of the simulation. It represents the actual PV as measured by the measurement device (transmitter) before the DCS or controller actually samples the measurement.
• PV in Digital Form - This is calculated from the PV in Analog Form and is numerically very close to it. It is first sampled from the PV in Analog Form and subject to the Hardware Resolution and then the PV Filter Time Constant.
• Output in Digital Form - This is the controller output in the electronic controller (usually a DCS) before output from the computer. If in auto, it is calculated from the PV in Digital Form, the setpoint and various settings including the tuning.
• Output in Analog Form - This is calculated from the Output in Digital Form. The calculation depends on the selected Final Control Element (pneumatic valve, Variable Speed Drive ("VSD"), or Electric Motor device) and is passed through a Digital-to-Analog (D/A) converter to become the Output in Analog Form and is affected by the Hardware Resolution setting.
• Actual Movement of Valve - This is calculated from the Output in Analog Form. The result depends on all of the Final Control Element settings. If you are simulating valve Stiction (sticking and then slipping) and Backlash (movement hysteresis or “play”), you will see a difference between this variable and “Output in Analog Form”.
• Disturbance - If you are simulating a Disturbance from outside the controller (e.g. something on the same process line affecting the flow or pressure), it appears here.
• Effect of Disturbance on PV - If you are simulating a Disturbance from outside the controller, this is its effect on the Analog PV.
• Feed-Forward Output - if you are simulating Feed-Forward control, this is the portion of the Output in Digital Form that results from the Feed-Forward settings.

Normally, you would note or capture the Absolute Error and the various Stats (Maximum and Minimum are shown) and then adjust the tuning to attempt to improve the result.




Simulator simulates the entire “environment” of the controller including sampling, A/D, filtering, output, D/A, valve stiction and backlash (or output to a VSD or electric device), disturbances and feed-forward. Simulator is complete and more importantly, realistic. Realism means that it will reduce time to trial various settings and process conditions and will also reduce the risk to the process..





























Here is an example of simulating valve stiction (stick and then slip). The controller outputs the red trace but the actual movement is the blue trace due to the stiction. You can simulate how different tuning settings might reduce the effects of this stiction or how much the process variation could be reduced by fixing the valve.

This image from the Standard mode. See the Advanced mode further below. You can also simulate a gain schedule based on the output (also called a Gain Output Characterizer) by entering relative gains below or above 1.00.




































The Simulator Controller tab is shown at left. Here you specify your controller’s PID equation form and your tuning settings are automatically converted to the ISA form for reference only. You can enter various controller options like PV filtering, deadband and of course the control direction. In the Advanced mode, you can specify your controller’s A/D (Input) and D/A (Output) resolution.











This is the list of “PID Form"s. There is no need to check what specific model of controller you have and search through a database that is never up-to-date. You just need to know its form. This is much easier than you might think. The Simulator Help shows what most major manufacturers use and few controls use Derivative which then eliminates half the possibilities.




















The Simulator General tab is shown at left. Here you can specify initial conditions, the simulation duration and “What To Simulate”.













There are many options of “What To Simulate”. You can simulate constant setpoints if your control’s setpoint is rarely changed or setpoints that vary in various shapes or with a single step or multiple steps if your control’s setpoint is changed routinely by operators or frequently as a slave loop. Whether a control needs to primarily respond to setpoint changes (a slave control such as a flow control whose master is a level control or a blending control) or to disturbances is an important aspect of process control and control testing and optimization. You can even import a captured data stream of real-world setpoint variation from your DCS (provided the DCS can record and export it). You can simulate random variation and then reuse the identical (frozen) same random variation with different tuning settings. You can add noise to your setpoint. You can instead do all of these things to the Output in manual. You can use manual to prove that your Process settings match the real world, for example.














The Simulator Disturbance tab is shown at left. A disturbance is an upset or variation in the PV that is not caused by the controller. An example would be flow control where there is another user of that flow downstream that opens or closes its valve. There could be multiple disturbances in a single simulation. You can specify what kind of disturbance (“Disturbance Type”) (or none) occurs at what time and with what size. You can add noise to the Disturbance, and choose how much it effects the PV and whether it happens immediately or over some time (use Lag Time 1).











There are many options of Disturbance Type. If your control’s main function is to respond to setpoint changes, you may want to choose “No Disturbance”. The job of most controls is to compensate for disturbances. You can simulate disturbances of various shapes or with a single step or multiple steps. You can even import a captured data stream of real-world disturbance variation from your DCS (provided the DCS can record and export it). You can simulate random variation and then reuse the identical (frozen) same random variation with different tuning settings. You can add noise to the disturbance. A Disturbance allows you to verify the control with real-world conditions.






In general, in Process Control, if you can measure a disturbance to a control (e.g. there is usually a flow transmitter for steam flowing through a desuperheater and this is the primary disturbance) and if the disturbance does not appear later in the disturbance measurement than in the controller’s measurement, it may be possible to reduce the upset caused by the disturbance using Feed-Forward (“FF”) control. Most DCS’s PID blocks include this function. FF moves the controller’s output in proportion to the change in the Disturbance variable using a gain you specify (can be in the same or the opposite direction). In most cases, you need the correction to happen as soon as possible in which case, you only specify a gain both in the actual controller and in Simulator. In some cases, it needs to be delayed or filtered (smoothed) to emulate how it affects the controller’s PV later than the controller’s output does. You can enter these values directly in Simulator. In a controller, you would need to add logic external to the PID to do this. In most cases, only the gain is used. Simulator allows you to test the outcome of Feed-Forward under controlled conditions without risking your process or waiting for the right upset condition.






This is the complete “Final Control Element” tab in Advanced mode. The Final Control Element (“FCE”) is the final destination of the controller’s output, typically a valve, a VSD or an electric motor. Advanced mode includes additional settings for advanced users plus the Electric Motor Drive. Certain FCE’s are on/off electric motor devices such as electric control valves, headbox slice movement motors, stock refiner plate motors etc. They behave differently than pneumatic valves or VSD’s because the time to run the motor output, whose speed is usually fixed, is proportional to the size of the movement and is often significant. (The same is true for VSD’s but usually that time is not significant compared to the typical output changes). Also, electric motor drives often “coast” when the power is removed or sometimes there is a consistent amount of time lost to open relays and start the motor. These devices need to be simulated differently - otherwise you will have the wrong result!










This Simulator example illustrates the above point. A single 5 % manual output bump test is done (red trace). Since the Full Travel Time is 180 sec, this requires 9 sec to execute. The Process Time Constant is only 3 sec with zero dead time yet the PV (green) requires far longer to complete its change and settle out because the electric motor device requires this 9 sec to complete its move. It also coasts an additional 0.10 % so the Actual Movement of Valve exceeds the Output in Digital Form by 0.10 %. This is a very realistic simulation and produces a very different result than a VSD or pneumatic valve would.




































Too many settings or you got confused? Just “Restore All Defaults”.














Like with many of Analyse-Plus graphs, you can easily copy the created data to the clipboard (numbers as text) or to an Analyse-Plus native .csd file. You can then use all the graphing and analysis tools using that data or export it.





























Like all Analyse-Plus graphs, you can easily make screen captures in your favourite formats or copy to the Clipboard. The Settings page is saved/printed as well as the main graphing page in separate png’s. Filenames are even suggested based on the Title you enter at left.















We are process control experts and have been using our own product for almost 20 years! We are confident in its usefulness and realism.