HVAC
VRF and VRV system commissioning and startup field guide
Braze under nitrogen, pull a deep vacuum, weigh in the calculated charge, auto-address the controls, run the test mode, and file the startup record the warranty needs.
Direct answer
VRF (variable refrigerant flow), also called VRV, runs one or more variable-speed inverter compressors feeding many indoor units through a shared refrigerant network, modulating capacity to each zone. Commissioning it is a documented procedure, not a startup switch, because the long piping, the weighed-in charge, the deep vacuum, and the auto-addressed controls all have to be proven.
Key takeaways
- VRF is charged by weigh-in on a calibrated scale: factory charge plus additional charge calculated from actual liquid-line length and diameter, never trimmed by gauge pressure.
- Braze VRF under a low flow of dry nitrogen (a couple of psi) with the equipment unpowered so EEVs stay open, preventing oxide scale that fouls expansion valves.
- Evacuate VRF to 500 microns or below per the model manual, then run a decay test: isolate the pump and confirm the vacuum holds.
- ASHRAE 15 caps releasable charge divided by the smallest occupied room volume under the refrigerant RCL: about 26 lb/1000 cu ft for R-410A, near 4.6 for R-454B, 4.8 for R-32.
- The manufacturer startup record with real numbers is the warranty condition on most VRF lines: a blank or guessed entry is treated as no commissioning when a claim is reviewed.
What VRF is, and why commissioning it is not a startup switch
VRF, variable refrigerant flow, is a system that runs one or more variable-speed inverter compressors in an outdoor unit and feeds many indoor units through a shared refrigerant pipe network, varying the flow of refrigerant to each unit so every zone gets only the capacity it needs at that moment. VRV, variable refrigerant volume, is the same technology under Daikin's trademark. The two names describe one idea: modulate the refrigerant, not cycle the compressor on and off.
What makes commissioning different from a split system is that nearly everything is field-built and field-proven. The piping runs are long and branched, the charge is calculated and weighed in rather than factory-set, and dozens of indoor units have to find their address and report to the controller before the system knows what it is. A VRF is not commissioned by flipping the disconnect and feeling for cold air. It is commissioned by a sequence that proves the pipe is clean and tight, the vacuum was deep, the charge is right to the gram, every indoor unit answers, and the manufacturer's test run cleared with no faults.
The cost of skipping the sequence is not a callback next week. It is a compressor or a row of fouled expansion valves a year out, on a system where the manufacturer's warranty turns on a commissioning record that was never filled in. The work below is mostly about earning that record honestly.
What is the difference between heat pump and heat recovery VRF?
A heat pump VRF runs on a 2-pipe network and puts every indoor unit in the same mode: all heating or all cooling at once, never both. A heat recovery VRF can heat some zones while cooling others at the same time off a single outdoor unit, moving the heat it pulls out of one zone into another that needs it. That ability to run mixed modes is the whole reason heat recovery costs more and pipes differently.
The classic heat recovery layout uses three pipes, with the phase separation happening at the outdoor unit and hot gas and liquid distributed to branch controllers that decide, per zone, whether to deliver heating or cooling. Some manufacturers achieve the same simultaneous operation on a 2-pipe network by doing the phase separation inside a branch controller (a BC box) instead of the outdoor unit, so confirm the actual architecture against the equipment, not the label on the box. Either way, the branch controller is the part that makes simultaneous heat and cool possible, and it is a commissioned component with its own valves and sensors.
Pick the wrong one for the building and you find out at commissioning. A heat pump system in a building with a perimeter that needs heat while the core needs cooling will fight itself all shoulder season, and no amount of balancing fixes a mode limitation that was decided at design.
The refrigerant piping network and its limits
The piping is the part of a VRF that least resembles a split system. Refrigerant leaves the outdoor unit, runs a long main, and splits at the manufacturer's branch fittings, the refnet joints and distribution headers, to feed each indoor unit or branch controller. Those branch fittings are engineered for the phase and flow at that point in the network. They are not generic copper tees, and substituting a hardware-store tee is one of the fastest ways to wreck oil return and refrigerant distribution.
Every line is sized by the manufacturer's design tool, not by what fits the nearest fitting. The size depends on capacity downstream of that point, and it changes as the network branches. The total network length, the longest single run from outdoor to the farthest indoor unit, and the height differences all have ceilings the manufacturer publishes, and exceeding them starves units and traps oil.
Carry the limits as ranges and verify the exact number against the design manual for the specific model, because they vary by manufacturer and system size. A single outdoor-to-indoor run often tops out somewhere around 120 to 165 m of actual length, total system equivalent length runs into the hundreds of meters, and the outdoor-to-indoor height difference is commonly capped near 50 m, with indoor-to-indoor height on a branch held tighter, often 15 to 30 m. These are oil-return and distribution limits, not suggestions, and a design that quietly busts one of them shows up as a unit that will not heat in a far corner.
| Piping limit | Common range (confirm per manufacturer) |
|---|---|
| Longest actual outdoor-to-indoor run | About 120 to 165 m |
| Total system equivalent length | Into the hundreds of meters |
| Outdoor-to-indoor height difference | Commonly up to about 50 m |
| Indoor-to-indoor height on a branch | Often 15 to 30 m |
| Branch fittings | Manufacturer refnet joints and headers only |
Why braze a VRF under a flowing nitrogen purge?
You braze a VRF under a low flow of dry nitrogen so the inside of the copper does not oxidize. When you heat copper in air, the oxygen in the pipe reacts with the hot copper and forms a layer of black copper oxide on the inside wall. As the joint cools, that oxide flakes off into loose scale, and the scale travels with the refrigerant and oil straight into the smallest passages in the system: the electronic expansion valves at every indoor unit and the outdoor unit's own valves. A fouled EEV is the number-one self-inflicted VRF killer, and it traces straight back to brazing dirty.
The technique is simple and not negotiable. Flow dry nitrogen through the line at a trickle, just a couple of psi, so it displaces the air and sweeps the inside of the joint while you braze. Too much flow blows the molten filler; too little lets oxygen back in. You purge from one open end and let the nitrogen out the other, never into a dead-ended, sealed pipe that can build pressure.
One trap is specific to VRF and catches people who have only done splits. Do not energize the equipment before the brazing and purge are done. Powering up can drive the EEVs closed, and a closed EEV blocks the nitrogen path, so your purge stops at the valve and the pipe past it oxidizes anyway. Braze it clean first, power it later.
Pressure test, evacuation, and the decay hold
After the pipe is brazed it gets pressure tested with dry nitrogen, not refrigerant, and not shop air. You bring the system up to the design test pressure the manufacturer specifies for that refrigerant and hold it for a long period, commonly a full day, watching for a pressure drop that temperature change alone cannot explain. Any unaccounted drop is a leak, and you find it and fix it before you go further. A VRF has dozens of brazed joints and flare connections, so one missed joint on a long run is the realistic failure, not a bad coil.
Then you evacuate, and a VRF wants a deeper vacuum than a residential split because the internal volume is large and the runs are long. The job is to pull out air, nitrogen, and especially moisture. Many manufacturers call for evacuation to 500 microns or below, and some specify deeper for large systems, so use the number in the commissioning manual for that model. A triple evacuation, pulling down, breaking the vacuum with dry nitrogen, and pulling again, is the practical way to scrub stubborn moisture out of a big system before the final pull.
The decay test is what proves it. Once you hit the target, isolate the vacuum pump, close the valve, and watch the micron gauge. A vacuum that holds is dry and tight. A vacuum that climbs slowly and levels off is still outgassing moisture and needs more pump time. A vacuum that climbs and keeps climbing is a leak you have to chase. Moisture and non-condensables left in a VRF do not stay quiet. They form acid with the POE oil, freeze at the expansion valves, and corrode the system from the inside, and they are the reason the deep vacuum is the step nobody is allowed to rush.
How is a VRF system charged?
A VRF is charged by calculation and weighed in, not trimmed by gauge pressure. The outdoor unit ships with a factory charge sized for a defined liquid-line length, often somewhere around 30 to 50 m depending on the model. Any pipe beyond that holds refrigerant the factory charge does not cover, so you calculate an additional charge from the actual liquid-line length and diameter, look it up against the manufacturer's per-foot or per-meter table, total it, and weigh that exact amount into the system on a calibrated scale.
Diameter drives the per-length amount hard, because volume goes up with the square of the pipe. A small liquid line might hold a fraction of an ounce per foot while a large one holds several ounces per foot, so a network with mixed line sizes is added up segment by segment, not by total footage at one rate. The charge is added in liquid phase through the liquid service port, weighed, and both the factory charge and the calculated additional charge are recorded on the startup sheet.
Use the manufacturer's table for the actual refrigerant. An R-454B system does not take the same per-length charge as an R-410A system, because the densities differ, and one manufacturer's table does not transfer to another's. Many systems also have an automatic charging or charge-check mode that the outdoor unit runs to verify the amount once the network is pressurized and operating, and that check is part of commissioning, not a substitute for weighing in the calculated value. For how superheat and subcooling read on a metering-device system once it is running, see the refrigerant charging field guide; on a VRF the weigh-in is the primary method and those readings are a cross-check, not the charge target.
| Charge input | Where it comes from |
|---|---|
| Factory charge | Outdoor unit nameplate, covers a set liquid-line length |
| Additional charge | Liquid-line length and diameter, manufacturer table |
| Refrigerant type | R-410A or A2L (R-32, R-454B), changes the per-length rate |
| Method | Weighed in liquid through the liquid port on a calibrated scale |
| Verification | Auto charge-check mode plus recorded weighed amount |
Controls wiring, comms, and auto-addressing
The control side of a VRF is a daisy-chained communication bus, not a thermostat per unit. The outdoor unit, the indoor units, the branch controllers, and the central controller all sit on a shared comms wire, and the system has to map who is where before it can run. Get the wiring right first: the comms is typically run in a continuous loop or daisy chain following the manufacturer's topology, polarity matters on many systems, and the comms is kept separate from the line-voltage power so noise does not corrupt the bus.
Auto-addressing is the step where the outdoor unit polls the bus and assigns each indoor unit an address, then reports how many it found. When the count it finds does not match the count you installed, you have a problem to chase before anything else: a unit with no power, a broken comms leg, a wiring polarity error, or two units someone manually set to the same address. Address conflicts are common when units were pre-set in a shop and then mixed on site, and they will stop commissioning cold.
The tell that addressing is healthy is a clean count and no communication faults on the controller. If the system reports fewer units than exist, walk the chain from the outdoor unit and find the break, because a unit that never addressed will never run, never report a fault, and never show up in the test run. It is simply invisible until the owner notices that one room is dead.
The test run and commissioning mode
Every major VRF line has a manufacturer test run, a commissioning routine the outdoor unit drives that exercises the whole system and reports what is wrong. It checks that every indoor unit is addressed and communicating, cycles the electronic expansion valves and confirms they step, reads the sensors and flags any that are open or shorted, verifies the comms to the branch controllers, and confirms the system can build the right pressures in heating and cooling. The output is a pass or a list of error codes.
Run it, and clear every code before you call the system commissioned. This is the part that separates a real startup from a checkbox. An EEV that fails to step shows up here, and a healthy valve coil reads a consistent resistance across all units, so an outlier coil is a flag to investigate. A swapped sensor, a stop valve someone left closed, a comms fault on one branch, all of it surfaces in the test run if you actually read the results instead of waiting for the unit to stop beeping.
Do not interpret the absence of an alarm as a pass. Confirm the test run completed and reported a clean result, with the expected number of indoor units, in both heating and cooling where the season allows. A VRF will happily run on a partial system and feel fine in the room you are standing in while a far branch never came online. The test run is how you find that out on commissioning day instead of in January.
Does ASHRAE 15 limit how much refrigerant goes in a room?
Yes. ASHRAE Standard 15, the refrigerant safety standard, sets a refrigerant concentration limit, the RCL, which is the most refrigerant that may be present in an occupied space if the whole charge serving that space leaked into it. The check is the releasable charge in the circuit divided by the volume of the smallest occupied room that circuit serves, and that number has to stay under the refrigerant's RCL. A VRF puts a large charge into a long network feeding many rooms, so the smallest room on the system is the one that governs.
The RCL is expressed in pounds per 1000 cubic feet, and the values are far lower for the A2L refrigerants than for what they replace. R-410A carries an RCL around 26 lb per 1000 cubic feet, while its A2L replacement R-454B is far lower, in the range of about 4.6, and R-32 lands near 4.8. The exact value and the calculation method come from the current edition of the standard and its addenda, so run the real number, do not carry one from memory, and confirm the edition the jurisdiction adopted.
When the releasable charge exceeds the RCL for the smallest room, the standard calls for mitigation: refrigerant detection in that space, alarms, ventilation that runs on a leak, and on many systems automatic shutoff valves that isolate the leaking section so the whole charge cannot dump into one room. This is not optional paperwork. A VRF that ignores the RCL can put a dangerous concentration into a small office or hotel room off the same circuit that serves a large open area, and the small room is exactly the one the design tends to forget.
A2L refrigerants in VRF
The phase-down of high-GWP refrigerants has moved VRF onto A2L refrigerants such as R-32 and R-454B, classified A2L under ASHRAE Standard 34: low toxicity, mild flammability. Mild is not zero. A2L will burn under the right concentration and ignition, which is why the mitigation that used to be a nicety on a large R-410A system is now built into the rules and the equipment for A2L.
On the job that means leak detection sensors in occupied spaces where the charge warrants them, controls that on a detected leak shut down the system and start ventilation, and shutoff valves that isolate sections so a single leak cannot release the whole network charge into one room. The flammability also changes how you handle the gas: keep ignition sources away during charging and service, watch the charge-per-room math harder than you would with an A1 refrigerant, and follow the manufacturer's A2L-specific install requirements, which are not the same as the older R-410A manuals.
The piping, brazing, vacuum, and weigh-in are the same discipline as before. What changed is the safety case around a leak, and on an A2L VRF the commissioning is not done until the leak detection, the mitigation ventilation, and the shutoff response have been tested and proven, not just installed.
Condensate drains, traps, and the overflow test
Every cooling indoor unit makes condensate, and a VRF can have dozens of them, often ducted units hidden above ceilings where a drain problem turns into a stained ceiling before anyone smells it. Each unit needs its drain pitched to fall, a trap sized for the unit's static so it seals against air being pulled or pushed through the drain, and, for the many units that sit below the drain line, a working condensate pump.
Slope is the cheap failure. A drain run that looks level to the eye but actually dips holds water, grows biofilm, and clogs, and then the pan overflows. Pitch it, support it so it stays pitched, and trap it per the unit's airflow direction, because a missing or wrong trap lets the blower either hold water in the pan or pull air through the drain and stop it from draining.
Test it before the ceiling closes. Pour water into each pan, confirm it runs out and the pump lifts it where a pump is fitted, and verify the overflow safety switch actually cuts the unit when the pan backs up. The float switch that nobody tested is the one that fails to trip on the day the drain clogs, and the first sign of that is water coming through the ceiling tile of the office below.
Power, the inverter, and grounding
The outdoor unit is an inverter drive, and that changes how it draws power. Inverter compressors soft-start, ramping speed up instead of slamming across the line, so the brutal locked-rotor inrush of an old across-the-line compressor is largely gone. That does not mean you size the circuit by feel. Size the breaker and conductors to the manufacturer's minimum circuit ampacity and maximum overcurrent protection on the nameplate, because those numbers account for the drive, and an oversized breaker defeats the protection the manufacturer designed in.
Voltage at the unit matters more on a long feeder than people expect, and inverter electronics have a tighter low-voltage tolerance than an old contactor-and-relay unit. A rooftop or far-corner outdoor unit at the end of a long run can sit low enough to throw a voltage fault, so check the run length against the drop the same way you would for any sensitive load. For sizing the feeder and holding voltage drop in range, the voltage drop field guide covers the calculation; the VRF point is to verify it, not assume it.
Grounding is not where you economize on a drive. Inverter outdoor units generate electrical noise and need a solid equipment ground both for fault clearing and for the comms bus to stay clean. A poor ground shows up as nuisance comms faults that look like a control problem and are really a grounding problem.
Airflow, filters, and external static on ducted heads
A VRF gets the refrigerant right and still underperforms if the air side was never checked. Ductless cassettes and wall heads need their filters in and clear and their throw pattern unobstructed, which sounds obvious until you find the protective film still on a cassette grille at startup. The refrigerant can be perfect and the room still will not condition if the air is not moving the way the unit needs.
Ducted indoor units are where airflow gets missed most, because they hide a fan and a small amount of external static capability above a ceiling. Each ducted unit has a rated external static pressure, and ductwork that exceeds it starves the coil, drops capacity, and on heat-recovery systems confuses the unit about what it is doing. Set the ductwork and the unit's fan setting so the measured airflow matches the design, and treat a ducted VRF head like any other air handler that has to hit a number.
Where the project includes a balance, coordinate it with the commissioning so the air side and the refrigerant side are proven together. For the air-side test, adjust, and balance sequence and the report that documents it, see the air balancing field guide. A VRF commissioned with the refrigerant verified and the airflow never measured is half-commissioned, and the half nobody checked is the one the occupant feels.
Central control, BMS integration, and owner training
Most VRF jobs end at a central controller and, on commercial work, a tie into the building management system. The central controller schedules the units, sets limits, and gives the owner one place to run the system. The BMS integration usually comes through a manufacturer gateway speaking BACnet or Modbus to the building controls, and that gateway has to be addressed, mapped, and tested like any other commissioned point, not just bolted on.
Prove the integration with the building actually controlling, not the gateway sitting on the bench. Confirm the schedules run, the setpoint limits hold, the system reports its status and faults up to the BMS, and a command from the front end actually moves a unit. A common miss is mode conflict on a heat pump system, where the BMS asks one zone for heat while the master is locked in cooling, and the unit quietly ignores it. Catch that at commissioning by exercising the modes through the BMS.
Then train the owner on the system they actually have. Show them how the central controller schedules, what a fault code means and who to call, how the mode limitation works if it is a heat pump system, and where the leak detection and shutoff live on an A2L system. An owner who does not understand the controls will defeat them within a month, and the comfort complaints that follow get blamed on the equipment that was working fine.
The startup record and the warranty
The manufacturer's commissioning sheet is the document that closes a VRF job, and on most lines the equipment warranty depends on it being completed and submitted. It is not a formality the office files and forgets. It is the evidence that the pipe was tested, the vacuum was deep, the charge was calculated and weighed, the units addressed, and the test run cleared, and it is what the manufacturer asks for when a compressor fails under warranty.
Fill it in as you go, with real numbers, not from memory once the crew has packed up. The piping lengths and the additional-charge calculation, the weighed charge, the vacuum level and the decay result, the test-run pass with the indoor-unit count, the RCL check and the leak-detection test on an A2L system, and the error log showing every code cleared: those are the entries that matter, and a blank or a guessed number on the sheet is the same as no commissioning at all when the claim is reviewed.
The honest version of this is short to say and easy to skip under schedule pressure. A VRF whose startup record is incomplete is a VRF whose warranty is at risk, and the day you need the warranty is the day the record gets read line by line.
What to document
The startup record reconciles the design to what was built and proves each commissioning step. Capture the network, not just the outdoor unit, because the charge and the limits all trace back to the pipe.
Record the system and refrigerant, the piping lengths that drove the charge, the factory plus calculated additional charge against the weighed amount, the vacuum level and decay result, the test-run outcome with the indoor-unit count, the ASHRAE 15 RCL check and the leak-detection test, and confirmation that every error code was cleared. The person reading this a year from now needs to reproduce the charge and confirm the system was proven, so write the inputs, not just the verdicts.
| Field to record | Why it matters |
|---|---|
| System type and refrigerant | Heat pump vs heat recovery, R-410A vs A2L, sets the rules |
| Liquid-line length and diameter | Drives the additional-charge calculation |
| Factory + additional charge vs weighed | Proves the charge was calculated and weighed in |
| Vacuum level and decay result | Shows the system was dry and tight |
| Test-run pass and indoor-unit count | Confirms every unit addressed and ran |
| RCL check and leak detection tested | ASHRAE 15 compliance for the smallest room |
| Error codes cleared | No open faults at handover; warranty evidence |
Common mistakes
- Brazing without a flowing nitrogen purge, leaving oxide scale that fouls the expansion valves.
- Estimating the charge or topping off by gauge pressure instead of weighing the calculated amount.
- Accepting a weak vacuum and skipping the decay hold, leaving moisture and non-condensables in the system.
- Two indoor units set to the same address, or a unit that never auto-addressed and stays invisible.
- Ignoring the ASHRAE 15 RCL for the smallest occupied room on the circuit, especially on an A2L system.
- Closing the ceiling before pouring water through each condensate pan and testing the overflow switch.
- Using generic copper tees instead of the manufacturer's branch fittings, wrecking oil return and distribution.
- Calling it done when the test run alarm is off, without confirming a clean pass and the full unit count.
Field checklist
Want this checklist to run itself on every job — with photo proof and a signed record crews can hand the customer? That's FieldOS.
Standards and references
The document that actually governs a VRF startup is the equipment manufacturer's installation and commissioning manual. It sets the design pipe lengths and height limits, the branch fittings, the test pressure, the vacuum target, the additional-charge table, the auto-address and test-run procedure, and the startup sheet the warranty needs. When anything below conflicts with that manual for the specific model, the manual wins.
ASHRAE Standard 15 is the refrigerant safety standard that sets the concentration limits and the mitigation requirements for the charge in occupied spaces, and ASHRAE Standard 34 is the classification standard that defines the safety groups, including the A2L group that R-32 and R-454B fall into. The exact RCL values and required addenda come from the current edition, so verify them rather than quoting from memory. AHRI Standard 1230 is the performance rating standard for VRF multi-split equipment, which is where the published capacity and efficiency ratings come from.
Refrigerant handling, recovery, and the certification to do it fall under EPA Section 608, which the refrigerant charging field guide covers in more detail. The mechanical code, the International Mechanical Code in much of the United States, carries the refrigeration and machinery-room provisions and adopts refrigerant safety requirements by reference. Codes are adopted and amended by jurisdiction, so confirm the adopted edition and any local amendments with the authority having jurisdiction before you treat a number as final.
Units, terms, and conversions
VRF spans two trademark names and a mix of imperial and metric units, since much of the equipment is designed in metric and installed against imperial tools.
VRF and VRV name the same technology. Pipe lengths and height limits are usually published in meters, while charge tables may give per-foot or per-meter amounts and charge in ounces, pounds, or kilograms. Vacuum is measured in microns of mercury, where lower is deeper and drier. RCL is given in pounds per 1000 cubic feet. Know which unit the manufacturer's table is in before you calculate, because mixing meters and feet in a charge calculation is a quiet way to land the wrong weight.
- VRF / VRV
- Variable refrigerant flow / variable refrigerant volume, the same modulating multi-split technology under different trademarks
- Heat recovery
- A VRF that can heat some zones and cool others at once, moving heat between them off one outdoor unit
- Branch controller
- The valved box (refnet or BC unit) that splits refrigerant and sets each zone's mode on a heat recovery system
- Additional charge
- Refrigerant added beyond the factory charge, calculated from liquid-line length and diameter and weighed in
- Micron
- A unit of vacuum; a lower micron reading is a deeper, drier vacuum, with VRF targets commonly 500 microns or below
- EEV
- Electronic expansion valve, the metering device at each indoor unit, the part oxide scale fouls first
- RCL
- Refrigerant concentration limit per ASHRAE 15, the most refrigerant allowed in the smallest occupied room if it leaks
- A2L
- An ASHRAE 34 safety class for low-toxicity, mildly flammable refrigerants such as R-32 and R-454B
FAQ
What is a VRF system?
A VRF (variable refrigerant flow) system uses variable-speed inverter compressors in an outdoor unit to feed many indoor units through a shared refrigerant network, varying flow to each so every zone gets only the capacity it needs. VRV is the same technology under Daikin's trademark.
What is the difference between heat pump and heat recovery VRF?
A heat pump VRF runs every indoor unit in one mode, all heating or all cooling, on a 2-pipe network. A heat recovery VRF heats some zones while cooling others at once, using branch controllers and usually a 3-pipe layout to move heat between zones off a single outdoor unit.
Why braze a VRF with nitrogen flowing through the pipe?
Heating copper in air forms oxide scale inside the pipe that flakes off and fouls the electronic expansion valves, the number-one VRF failure. Flowing dry nitrogen at a couple of psi displaces the oxygen so no scale forms. Keep the equipment unpowered so the EEVs stay open and the purge can flow through.
How is a VRF system charged?
A VRF is charged by weigh-in, not by gauge pressure. The factory charge covers a set liquid-line length; you calculate the additional charge from the actual liquid-line length and diameter against the manufacturer's table, then weigh that exact amount in on a calibrated scale. Record both the factory and calculated charge.
How deep does a VRF system need to be evacuated?
Deeper than a residential split, because of the large volume and long runs. Many manufacturers call for 500 microns or below, and some specify deeper for large systems, so use the model's commissioning manual. After hitting the target, run a decay test: isolate the pump and confirm the vacuum holds, proving the system is dry and tight.
Does ASHRAE 15 limit the refrigerant charge in a VRF?
Yes. ASHRAE 15 sets a refrigerant concentration limit (RCL): the releasable charge divided by the smallest occupied room's volume must stay under it. A2L refrigerants like R-454B have much lower RCLs than R-410A, so confirm the current edition's value and add leak detection and mitigation where the charge exceeds it.
What do I do if auto-addressing finds fewer indoor units than I installed?
Walk the comms chain from the outdoor unit and find the break before going further. A missing unit usually means no power, a broken or wrong-polarity comms leg, or two units set to the same address. A unit that never addressed stays invisible, never runs, and never reports a fault, so the count must match the install.
Why does my VRF have units that will not heat or cool in a far corner?
Most often the piping busted a manufacturer limit: too long a run, too much height difference, undersized line, or a generic tee instead of a branch fitting, all of which starve a far unit or trap oil. Check the run against the design limits, and confirm the unit addressed and passed the test run.
Do A2L refrigerants change how a VRF is commissioned?
The piping, vacuum, and weigh-in stay the same, but A2L refrigerants are mildly flammable, so commissioning adds proven safety steps. Test the leak detection, the mitigation ventilation, and any automatic shutoff valves, keep ignition sources away during charging, and follow the manufacturer's A2L-specific install manual, which differs from older R-410A documentation.
Why does the manufacturer startup sheet matter for a VRF?
On most VRF lines the equipment warranty depends on a completed commissioning record. It documents the pipe test, the deep vacuum, the calculated and weighed charge, the addressing, and the cleared test run. A blank or guessed entry is treated as no commissioning when a warranty claim is reviewed, so fill it in with real numbers.
People also ask
Codes cited in this guide
This guide is written and reviewed against the published standards below. Always confirm the current adopted edition with the authority having jurisdiction.