How to Check for Interaction Between Two Systems

Why this matters

Your system worked fine, then another trade installed something nearby, and now yours faults, but every part of yours tests good. This is the fault that lives between two systems, each correct on its own, in conflict only together. It is the class that makes a good tech look bad and turns into a finger-pointing match with the other contractor. A clean interaction test settles who owns what without the fight. The other install is not defective and neither is yours; the conflict is emergent, and you have to prove it, not argue it.

Start from what changed since it worked

The tell for this family is a system that ran right and broke after adjacent work went in. Establish that first: when did yours last work correctly, and what got installed, modified, or added nearby since? A new circuit, a new appliance, a remodel that changed airflow or drainage, a smart control another trade wired in. Because the new work is correct in isolation, nobody thinks to connect it to your fault.

Step 1: Run each system alone

Isolate them. Run your system with the other one fully off, and see whether the fault appears. Then run the other system with yours idle. A fault that shows only when both run, and never when either runs alone, is an interaction rather than a component failure. If yours faults even with the other system fully off and isolated, it is your fault to own, and the adjacent work is a coincidence, not a cause.

Step 2: Run them together and measure the shared node

When both run and the fault shows, do not measure the two systems, measure the resource they share, under combined load. The interaction almost always runs through one shared, finite path:

• A shared electrical circuit or service, where voltage sags under combined draw or a motor start
• A shared water supply, drain, or vent, where pressure or capacity gets split too thin
• A shared gas supply or combustion air
• A shared ground, bond, or signal reference, where crosstalk, noise, or a ground loop appears
• Shared airspace, a shared pressure envelope, or plain thermal proximity

The number at the shared node under combined demand tells the story a single-system test is built to hide.

Step 3: Confirm by removing the newer system

Take the adjacent system out of the picture. Unplug it, close its valve, kill its breaker, and watch whether your fault clears. If it clears with the other system out and returns when you restore it, you have proven the interaction without condemning a single component. That is the whole diagnosis, and it is repeatable in front of anyone who doubts it.

Step 4: Name the mechanism, then the fix

The fix relieves the interaction, it rarely condemns a part, because neither part is broken. Reduce one system's demand, stagger the two so they do not run at the same moment, isolate the coupling with a dedicated circuit or a separated supply or a proper bond, or raise the shared resource's ceiling. Which end you correct is a coordination decision, not a blame decision.

Step 5: Handle the other trade straight

Emergent conflict is nobody's defect, so do not frame it as the other contractor's mistake or as your own. Report it as what it is: two correct installs sharing a resource that was never sized for both together. That framing gets the other trade cooperating instead of defending, and your run-alone versus run-together results keep the conversation on evidence instead of opinion.

References

• Trade-standard practice for systems-level and shared-resource diagnosis
• Manufacturer minimum-supply specifications (voltage, pressure, gas, combustion air)
• Applicable sizing codes for conductors, piping, drains, and venting
• See related: The Interaction Between Two Loads That Creates a Fault; An Adjacent System Might Be Causing Your System's Fault