# From free scan to lasting calm.

> From the free PID Performance Scan through optimisation at a fixed price to Stay Optimal for a fixed annual fee. After every phase you decide again.

The programme consists of three services that build on each other. You always start with the free PID Performance Scan and decide again after every phase whether you want to continue. Every phase has a price you know in advance, and your plant simply keeps running the whole time.

## Phase 1: PID Performance Scan (free)

- **Introduction**: A short video call, no travel needed. We want to understand which control loops there are, where oscillations occur and where operators intervene. You do not need to prepare anything.
- **Collecting data**: A CSV export from your DCS or historian, supplemented with screenshots of the control loops. No extra sensors, no modifications. If an export is not possible, we work from the trend charts themselves.
- **Process analysis**: For each loop we derive how the process behaves: dead time, time constant and process gain. We also map oscillations, control activity and the margin to trip limits. That shows us where the loop becomes restless and why.

[More about the scan](/en/services/performance-scan)

**Decision point:** In a video call we discuss the findings with you or your engineer and you receive a fixed price for the optimisation. Together we also agree on the metrics we will measure before and after, so it is clear in advance when the optimisation has succeeded. Then you decide whether you want to continue. If there is little to improve, we will tell you honestly.

## Phase 2: Optimisation (fixed price)

- **Calculate**: Within ten working days we calculate the new settings: Kp, Ti and, where needed, Td, plus limits and the dead band. We discuss every change with you or your engineer in advance, including the reasoning behind it.
- **Implement**: Your own operator enters the settings step by step, we support on site. All during normal operation, with continuous monitoring of process behaviour and always a fallback to the old settings.
- **Verify**: Two to four weeks later we compare the control behaviour before and after on the same metrics, remotely, using fresh trend data. After completion and payment you receive the full report for your own records.

[More about the optimisation](/en/services/optimisation)

**Decision point:** Your process runs demonstrably more calmly. A periodic review is optional after that: you choose whether you want the control loops to be followed up.

## Phase 3: Stay Optimal (fixed annual fee)

- **Measure again**: Usually once a year we review every optimised loop again on the same metrics. On site, or remotely via an OPC connection to your DCS or historian.
- **Adjust where needed**: If a loop starts to drift, we first look for the cause and then make a substantiated adjustment, during normal operation. Where nothing needs to change, we change nothing.

[More about Stay Optimal](/en/services/stay-optimal)

## Result: A process that runs calmly, and stays calm.

Fewer trips, fewer alarms and less manual intervention. With a periodic review it stays that way, even as the process or the load changes.

## Among others, we work with these systems

Yokogawa, Honeywell, Siemens, Emerson, ABB, Rockwell Automation

## Whatever control loop you run

We analyse and optimise industrial control loops: from flow, pressure, temperature and level control to the interactions between several loops in one plant. When was your control loop last properly calculated, instead of adjusted by feel? Below is a selection of processes we have experience with. If yours is not listed, we are happy to think along with you.

[What a PID controller does and how to tell it is set wrong](/en/knowledge/pid-controller-explained)

- **Can overshoot as soon as the flow changes** (Temperature): When the flow in the process changes, the temperature can overshoot and stabilise with a delay. This can happen because thermal mass and/or transport delay sits between control action and measurement, depending on how well the loop is tuned for it.
- **Valve position can become nervous** (Flow): The valve can keep making tiny corrections, as if the loop cannot make up its mind. This can happen because flow loops react directly to every measured change, including measurement noise: changes that do not really exist in the process. That causes unnecessary wear and wasted energy.
- **Can quickly start to oscillate** (Level): The level can swing: as soon as the controller corrects too aggressively, the level shoots past the setpoint and then corrects back. This is typical of level because it is an integrating process. Without good control a tank simply runs empty or full, because there is no natural equilibrium to damp it.
- **Non-linear, with a tricky point around the setpoint** (Dosing): Dosing loops, for example for pH correction, behave non-linearly: around the equivalence point a small dose already causes a large change, while the same dose further away hardly does anything.
- **Loops that work against each other** (Interacting loops): Two loops that are each tuned perfectly well can still work against each other if they are coupled through the process: the correction of one loop disturbs the other, which in turn feeds back again. This often shows up as persistent unrest that does not go away, however often you retune one of the two on its own. We analyse control loops in relation to each other, not one by one.
- **Motor or pump can keep speeding up and slowing down** (Speed): The speed can fluctuate: you feel it in the plant and energy consumption rises unnecessarily. This happens because power does not rise linearly with speed: small speed variations cause disproportionately large power variations. Those fluctuations cost much more energy than you would expect.
- **Controller can keep overcorrecting** (Flue gas): The loop can keep correcting too hard or too late. This can happen because flue gas analysis adds considerable dead time to the loop through sample transport and analyser response time. The controller is therefore working with information from a few moments ago, reacts to it, but does not know that its previous action has already had an effect. The result: overcorrection.
- **Pressure fluctuation that is hard to stabilise** (Pressure): The pressure can keep fluctuating or gradually drift away. This can happen because pressure behaves very differently from system to system: in a closed vessel pressure is integrating (it keeps rising or falling without a natural equilibrium), in a pipe with throughflow it reacts more directly to the valve position.
- **Drum level moves against expectations** (Steam and boiler): As soon as the pressure changes suddenly, the drum level can react unexpectedly. When the pressure drops, part of the water turns into steam and the steam bubbles expand (swell): the level appears to rise, while no water has been added. When the pressure rises, the opposite happens (shrink). The controller sees these volumetric effects as real level and corrects the wrong way.
- **Product can vary, while the ratio seems right** (Ratio): Even though both flows show the correct ratio on screen, the end product can vary. This can happen when the ratio is not matched precisely enough to the actual process.
- **Behaviour can be counter-intuitive, controllers can disturb each other** (Cascade and feedforward): Cascade and feedforward structures can cause unexpected behaviour. In a cascade the inner loop can act so hard that the outer loop no longer sees a deviation to react to. In feedforward, a wrong gain on the disturbance variable can actually push the process further out of balance. These structures require tuning across the layers, not just tuning per layer.
- **Vacuum drifts away, and leaks make stability harder** (Vacuum): Vacuum is integrating: without a controller it drifts away on its own. At the same time even small leaks can introduce a persistent disturbance that is hard to control out. This requires careful tuning of the controller: aggressive enough to reach the setpoint, but stable enough to absorb leaks.

## Ready for a first conversation?

The programme starts with a no-obligation introduction by video call. After that, the free PID Performance Scan shows what there is to gain in your control loops and what it costs. Would you like to see what it delivered elsewhere first? Then take a look at the results.