- The worst time to decide who gets called, who approves spending, and what is protected first is during the emergency. A one-page plan written in advance saves hours.
- Not every failure is an emergency. Triage rules based on consequence (food, medication, data, safety, occupant health) keep after-hours calls for the things that justify them.
- Critical spaces need their own plans: what temperature or condition must be held, how long you have, what temporary measures exist, and who decides.
- Response time in a vendor agreement means little unless it is defined: time to answer, time to arrive, and time to restore are three different promises.
- Good documentation, taken during the event, supports insurance claims, food safety decisions, and root-cause repair afterward.
Why a written plan beats a good phone number
Most facilities have a contractor's number somewhere. Far fewer have a clear answer to the questions that actually decide how an emergency goes: who is allowed to call at 11 pm, who can authorize a repair beyond a stated limit, what has to be protected first, and who will be at the building to let technicians in. When those are unresolved, the first hour is lost to phone calls, and in refrigeration or data environments the first hour is often the one that matters.
A response plan does not need to be elaborate. It needs to be accurate, accessible to the people who will use it, and tested at least once. Two pages is plenty for most single buildings.
Triage: deciding what counts as an emergency
Define emergencies by consequence rather than by equipment. A warm office is an inconvenience on a Saturday night; a warm walk-in with a week of inventory is a loss that grows by the hour. Writing the thresholds down prevents both over-calling, which wears out vendor goodwill, and under-calling, which costs product and safety.
| Level | Typical conditions | Suggested response |
|---|---|---|
| Immediate | Gas odor or suspected carbon monoxide, smoke or burning smell from equipment, refrigerant release in an occupied space, active water leak into electrical or ceilings | Follow safety procedures first (evacuate, shut off, call emergency services as appropriate), then call the vendor |
| Urgent after-hours | Failed cooling in a food, medication, lab, or IT space; total heating or cooling loss in a space with vulnerable occupants; failed hot water at a food service operation | Call the on-call line; start critical-space protocol; authorize diagnosis |
| Same day | Loss of cooling or heating to a general area, ice machine down in a busy operation, partial system failure with backups running | Call during the next staffed window or on-call line if impact grows |
| Next business day | Noise, minor comfort complaints, non-critical equipment faults, intermittent alarms with stable conditions | Submit through normal service request with data collected |
Gas odors, suspected carbon monoxide, fire, and major refrigerant releases are matters for the building's emergency procedures and emergency services first. HVAC response follows once the area is safe.
Critical-space plans
Pick the five or six spaces in your portfolio where a cooling or heating failure has outsized consequences, and write a short plan for each. These commonly include walk-in coolers and freezers, medication refrigerators, server and telecom rooms, labs, humidity-sensitive storage, and spaces with vulnerable occupants.
- What must be held: the temperature range or condition, and the point at which contents are at risk. Use your own regulatory or manufacturer requirements, not guesses.
- How long you have: an honest estimate of time to risk with doors closed and nothing heroic done. A full, well-insulated walk-in buys more time than a half-full one with a door that does not seal.
- First actions for staff: keep doors closed, move or transfer contents, start monitoring, record temperatures at intervals.
- Who decides to move product or occupants: named by role, with a backup.
- Temporary options: portable cooling, rental units, backup refrigeration at a nearby site, or arrangements with a supplier. Know where you would get each and what connections the building has for them.
- Monitoring: alarms, data loggers, or remote sensors that notify the right people, plus a plan for who receives the alert at night.
For food-service operations, align this with your food-safety procedures and local health department expectations, which govern what to do with potentially affected product. Our commercial refrigeration guide and walk-in guide explain how refrigeration fails and how long different failures tend to take to matter.
Temporary cooling and heating: plan the connections now
Temporary equipment is only fast if the site is ready for it. Portable spot coolers, rental chillers, and generator-powered units need somewhere to go, a way to reach it, power that can support them, and a path for hoses, ducting, or condensate. Walk the site with that in mind and note the limits.
- Identify where temporary equipment could be staged and whether it blocks egress or loading.
- Confirm electrical capacity and available connection points, and whether a qualified electrician is needed to connect.
- Note chiller or hydronic isolation valves and tie-in points if a rental chiller could ever connect to your loop.
- For rooftop systems, record crane or access constraints in case a unit must be replaced.
- Keep equipment specifications handy: model, serial, voltage, refrigerant, and capacity, so a supplier can identify a match quickly.
- Decide whether you have a generator, whether it runs HVAC loads, and what happens when it is not available.
In peak summer weeks, rental inventory and lead times can tighten region-wide. Do not assume availability; ask your vendor how they source temporary equipment and what the realistic window looks like.
Documentation during and after the event
Notes taken during an emergency are worth far more than reconstructions the next week. Designate someone to keep a simple log: time of discovery, readings, who was called and when, who arrived and when, actions taken, and decisions about product or occupants. Photograph conditions, gauge readings, and equipment data plates.
- Temperature records for affected spaces, from logs or manual checks.
- Alarm history and any controller fault codes before they are cleared.
- Technician findings, including what was observed before repair, not only what was replaced.
- Photos of failed components, and a decision on whether to retain them.
- Cost authorizations and who approved them.
- Notes for your insurer or landlord, as relevant to the lease or policy.
After the event, ask the contractor for a root-cause statement. A failed contactor may be the item that was replaced at 2 am, but the next question is why it failed. Our root-cause diagnostics guide lays out how to ask for that, and the DaVinci Portal keeps the event in the asset's history for the next technician.
What to ask in a service agreement
Many agreements promise a response time, and the words hide three different commitments. Ask the vendor to define each and say how the commitment is measured.
| Commitment | What to clarify |
|---|---|
| Answer time | Is the on-call line answered by a person, and how fast? Who triages the call? |
| Arrival time | Measured from the call or from dispatch? Does it vary by day, hour, or location? What happens during region-wide heat events? |
| Restoration effort | Is there a commitment to stabilize the space (temporary measures) even when the permanent repair requires parts? |
| Priority tiers | Do contract customers or critical spaces receive priority, and how is that defined? |
| Parts and stocking | Which common parts are stocked on trucks or in the shop, and how are special orders handled? |
| Escalation | Who is called if the first technician cannot resolve it, and how quickly? |
| Reporting | What documentation do you receive after each call, and in what format? |
Be realistic with yourself, too. A vendor cannot promise anything meaningful about a building it has never seen. Priority service is most credible when the contractor already has your equipment documented and a maintenance relationship in place. See our maintenance plans and guide to choosing a commercial HVAC contractor.
Test, then keep it current
Run a tabletop exercise once a year: pick a scenario, such as a walk-in failure on a holiday weekend, and walk through the plan with the people who would act on it. Check that phone numbers work, that someone has the building access codes, and that the person authorized to approve spending is reachable. Update the plan after any staffing change, equipment replacement, or lease change. A plan that names someone who left the company is worse than no plan because it feels reassuring.
Frequently asked questions
What should be in a one-page emergency plan?
Emergency contacts and their roles, triage definitions, a list of critical spaces with their limits, building access instructions, spending authority, and the location of equipment information. Keep it where on-call staff can reach it at night.
How do I decide whether to call after hours?
Use consequence. If waiting until morning risks food, medication, data, safety, or vulnerable occupants, call now. If the impact is comfort or convenience, collect details and submit during business hours.
Should I own backup equipment?
It depends on the space and the cost of failure. Some operations keep spare monitoring and small portable cooling on site; others rely on rentals. The key is knowing in advance which you will use and how it connects.
What does a guaranteed response time actually guarantee?
Only what the agreement defines. Ask whether it covers answering, arriving, or restoring, how it is measured, and what exceptions apply during widespread outages or weather events.
Who should keep records of an emergency?
Assign one person on site and ask the contractor for a written service report. Combined, those records support insurance, food safety, and later root-cause repair.
Have a site this applies to?
Davinci Mechanical is the commercial and union division of Scottish Tom's Heating & Air. Send us the equipment list or the problem and we'll tell you what we'd check first.