Decay ratio is a yardstick for reading how a loop settles after it is disturbed. When a well-behaved loop is bumped, it overshoots, swings back, overshoots less, and rings down to rest, and the decay ratio measures how fast that ringing dies out. The classic tuning target of one-quarter, meaning each swing is a quarter the size of the last, was for decades the default aim, but many modern practitioners now consider it too aggressive. This guide defines decay ratio, explains the quarter-amplitude criterion and the debate around it, and shows how to use it to read a setpoint-change response trend.
Decay Ratio / Quarter-Amplitude in one line: Decay ratio is the ratio of one overshoot peak to the previous overshoot peak in a loop's oscillatory response, describing how quickly the oscillation dies away. The classic quarter-amplitude criterion aims for a decay ratio of one-quarter, so each successive peak is a quarter the height of the one before. It was long a standard tuning target, but many practitioners now consider it too oscillatory and prefer smoother, more damped responses such as those from lambda tuning.
When you change the setpoint of a typical loop or hit it with a disturbance, the measurement rarely slides straight to the new value. Instead it overshoots, comes back and undershoots, overshoots again by less, and so on, ringing down in a decaying oscillation until it settles. Decay ratio quantifies that ring-down. It is the size of one peak divided by the size of the peak before it, measured from the final settled value. A decay ratio of one-half means each swing is half the previous one; one-quarter means each is a quarter; a very small ratio means the oscillation dies almost immediately.
This makes decay ratio a fast, practical way to read a response trend and judge how a loop is tuned. A loop whose successive peaks barely shrink, a decay ratio close to one, is lightly damped and dangerously close to sustained oscillation; it is tuned too aggressively and will ring for a long time after every upset. A loop whose overshoot vanishes after one small rebound is well damped and calm, perhaps even sluggish. By eyeballing just two consecutive peaks on a trend, an engineer can estimate the decay ratio and immediately characterize whether the loop is nervous, comfortable, or lazy.
Decay ratio pairs naturally with a couple of companion readings from the same trend, such as the overshoot on the first peak and the period of the oscillation. Together they paint a picture of the closed-loop behavior. But decay ratio is the headline number because it captures the crucial question of stability margin in a single ratio: how much closer to instability is this loop with each cycle, and how quickly does it recover?
The quarter-amplitude criterion, aiming for a decay ratio of one-quarter, has a long pedigree. It emerged alongside the earliest systematic tuning methods and became a default target because it represents a seemingly reasonable compromise: the loop responds quickly and each oscillation is markedly smaller than the last, so the response looks brisk and settles in a few cycles. For generations of engineers, tuning a loop to quarter-amplitude decay was simply what good tuning meant, and many classic tuning rules were designed to land near that target.
Modern practice has grown skeptical, and for concrete reasons. A quarter-amplitude response still overshoots substantially and rings for several cycles before settling, which means the loop is fairly lightly damped and has a modest margin to instability. In a plant where the process changes over time, through fouling, load shifts, or valve wear, a loop tuned to quarter-amplitude when conditions were nominal can slide into much heavier oscillation or outright instability as conditions drift. The aggressive edge that made it look responsive is the same edge that makes it fragile. Many practitioners now regard it as too oscillatory for comfort, a target that prioritizes apparent speed over robustness.
The contrast that captures the shift is with lambda tuning, which deliberately aims for a smooth, well-damped response with little or no overshoot and no ringing at all, trading a slightly slower approach to setpoint for a large stability margin and predictable behavior. Where quarter-amplitude accepts oscillation as the price of speed, lambda tuning treats a calm, non-oscillatory response as the goal. Neither is universally right; a loop that must reject disturbances fast may want more aggression, while one that must stay robust as conditions drift wants the smoother target. The value of understanding decay ratio is that it lets an engineer name and choose where on that spectrum a given loop should sit, rather than accepting one criterion by habit.
Decay ratio is read from the shape of a response trend, and trends are exactly what a SCADA historian stores. Every setpoint change and disturbance a loop experiences leaves a recorded response in the archive, so the decay of its oscillations can be assessed from data already captured, without staging a special test. A monitoring layer can scan these recorded responses and estimate decay ratio and overshoot automatically, flagging loops that ring for too long or whose oscillations barely decay.
In a cloud SCADA platform such as Merobix, this lets an engineer judge tuning quality across many loops and sites from the recorded behavior, rather than watching each loop live. A loop that once settled cleanly but now shows a decay ratio creeping toward one is a warning that the process has drifted or the tuning has aged, and the trend history makes that change visible over weeks. Because the analysis works from stored responses, it reaches loops on remote assets that an engineer would otherwise never see respond.
This serves any operation where loops must stay well-damped despite changing conditions. Oil and gas facilities, water and wastewater plants, power generation, and manufacturing all run loops whose response character matters, and being able to read decay from historized trends turns tuning assessment into something a central team can do at scale. It also grounds the aggressive-versus-robust tuning decision in real recorded behavior, so the choice of where to sit on the decay-ratio spectrum is informed by how each loop actually settles in service.
Decay ratio is the ratio of one overshoot peak to the previous one in a loop's oscillatory response, measured from the settled value. It describes how fast the ringing dies out after a disturbance or setpoint change. A ratio near one means the oscillation barely decays and the loop is lightly damped; a small ratio means it settles quickly.
Quarter-amplitude damping is a tuning target where the decay ratio is one-quarter, so each overshoot peak is a quarter the size of the previous one. It was long a standard aim because the response looks brisk and settles in a few cycles. Critics point out that it still overshoots substantially and leaves a modest stability margin, so a loop tuned that way can drift into heavier oscillation as process conditions change, which is why many now prefer smoother targets.
They sit at opposite ends of a tradeoff. Quarter-amplitude tuning accepts noticeable overshoot and several cycles of ringing in exchange for a fast-looking response. Lambda tuning deliberately targets a smooth, well-damped response with little or no overshoot and a large stability margin, trading some speed for robustness. Understanding decay ratio lets an engineer choose where a loop should fall between the two rather than defaulting to one.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.