Two alarms. One fires when you're running low. One fires when something feels dangerous. They can look identical from the outside — and require completely different responses.
Most of what we call "being dysregulated" or "struggling" or "activated" is not one thing. It is two very different system states that can produce similar-looking behavior and similar-feeling experience — but that emerge from entirely different mechanisms, require entirely different responses, and will get worse with the wrong intervention.
Understanding the difference is one of the most practically useful things this framework offers. Not just for clinicians matching treatment to presentation, but for anyone trying to understand why something they tried to help themselves or someone they love made things worse instead of better.
Think of the tank as the nervous system's available regulatory capacity at any given moment. It includes physical energy, attentional resources, the ability to tolerate uncertainty and ambiguity, and the physiological buffering that allows the system to absorb stress without destabilizing. It is not a fixed quantity. It depletes under demand — cognitive load, sensory input, emotional labor, social performance, unresolved pressure — and it refills with rest, safety, nourishment, co-regulation, and predictability.
When the tank is genuinely full and the threat detection system is quiet, something interesting happens: surplus capacity tends to emerge as exploration, creativity, play, and learning. This overflow is not a luxury. It is what the system does when it is actually free to update — to form new priors, to revise its models of the world, to develop flexible strategies. A full tank is the condition under which genuine growth becomes possible.
The tank does not refill through effort. It refills through safety, rest, co-regulation, and the reduction of unnecessary load. Trying harder when the tank is low is like trying to drive faster when the fuel gauge reads empty.
The burners are the nervous system's constraint detection infrastructure. They are not simply alarm bells. They are sophisticated monitors tracking whether the organism's core constraints — safety, belonging, autonomy, predictability — are under threat. When any of these are compromised, the burners activate. They signal that something requires a response. They generate arousal, urgency, and a drive toward action.
Burner activation is not pathology. It is the system doing exactly what it is designed to do. The question is what is triggering it, how sensitive the threshold has become, and whether the strategies the system selects to respond to it are working — or reinforcing the very condition that generated the activation in the first place.
Here is the crucial distinction. There are two types of burner activation, and they require opposite responses:
Trying to help someone think their way out of a capacity crisis doesn't work — the tank doesn't fill through reasoning. Trying to accommodate someone out of a threat-based pattern can maintain the very signal it's trying to lower — the prior that generated the threat response is never contradicted. Matching the response to the wrong alarm makes things worse.
Both types of activation produce pressure. Both produce dysregulation. Both can produce withdrawal, meltdown, shutdown, or acute distress. From the outside — and often from the inside — they can be indistinguishable. A child who refuses a transition might be running on a completely depleted tank and literally unable to process one more demand. Or they might be encountering a predictability threat — the routine changed and the system can't orient. These are different problems. "Let's talk about why this is happening" is useful in neither case but for different reasons.
The same person can be in different states at different times. A low-tank state that becomes chronic creates conditions for high-heat burners — a depleted system is more sensitive to threat because it has fewer resources to absorb uncertainty. This is why chronic stress and neurodivergent experience so often produce what looks like anxiety: the tank stays low from accumulated load, which keeps burner thresholds low, which means threat fires more easily, which further depletes the tank. The two alarms can become mutually reinforcing.
The most useful diagnostic question before any intervention is not "what is this person feeling?" It is: "which alarm is active?" The answer determines everything about what could actually help.
For a Low Water state, the primary intervention is restoration: reducing load, increasing safety, offering rest, providing co-regulation, removing unnecessary demands. The goal is to allow the tank to fill enough that the system can begin to operate with flexibility. There is no amount of insight, challenge, or exposure work that substitutes for this when capacity is genuinely depleted.
For a High Heat state, the intervention depends on what constraint is being tracked and whether the threat reading is calibrated to the current situation or driven by an older prior. Sometimes the environment actually is unsafe and the signal is accurate — the answer is genuine change in conditions. Sometimes a high-precision survival prior is being triggered by something in the present that pattern-matches to past danger — the answer is eventually some form of embodied disconfirmation, a new experience that contradicts the prediction the prior is running.
In both cases, the first step is the same: determine which alarm is active. Then proceed accordingly. The tank and the burners are different systems. They deserve different responses.