Commissioning a Replacement Condensing Unit
Purpose
This standing instruction guarantees that a new condensing unit leaves the property with its charge verified against a measured airflow baseline, its safeties proven, and every number written down against a target.
The failure it prevents is the one that reads correct. A tech charges to the superheat or subcool target on a system whose indoor airflow was never measured, gets a number inside the window, and signs it off. Low indoor airflow drops evaporator load, which drops suction pressure and drops superheat, so a tech chasing a superheat target on a starved coil removes refrigerant from a system that was already right. The gauges then agree with him. Three weeks later the unit is short on capacity, the compressor is running hot, and nobody can tell from the paperwork what was actually measured because nothing was.
Safety actions that gate the work
- Commissioning is energized work by definition, so the exposure is different from install. Meter with the covers on wherever the equipment provides test points, keep one hand out of the cabinet, and use a meter and leads rated for the circuit's category and voltage.
- The condenser fan restarts on a call with no warning. Nothing goes past the fan guard while the disconnect is closed, including a temperature probe on the discharge line.
- High-side pressures on current refrigerants exceed what a hose rated for older systems will hold. Use hoses and a manifold rated for the refrigerant in front of you, and stand out of line with any fitting you are cracking.
- If the unit is A2L (R-454B, R-32), no ignition source while any fitting is open, and the space stays ventilated during any charge adjustment.
- Recovered refrigerant during a trim goes to a recovery cylinder, never to atmosphere, at any charge size.
Scope
Covers startup and performance verification of a replacement air-cooled condensing unit or heat pump outdoor section on a residential or light-commercial split system, beginning at a signed pre-start checklist and ending at customer handover.
Does not cover the install itself, which is owned by the changeout day-of-install standard, and this procedure does not begin until that one's pre-start sheet is signed. Does not cover duct leakage measurement, owned by the duct leakage inspection SOP; airflow here is measured at the equipment, not at the ducts. Does not cover thermostat configuration beyond a staging check, owned by the thermostat replacement and configuration handover SOP.
Roles and handoffs
| Role | Owns | Hands off |
|---|---|---|
| Lead installer | The signed pre-start checklist and the physical install | Hands the commissioning tech a system ready to energize, with the line-set decision documented |
| Commissioning tech | Every step below, and the authority to stop the handover | Hands the office a completed startup sheet with values, not check marks |
| Office | Warranty registration and the inspection booking | Hands the customer the startup sheet copy |
| Service manager | Escalation when a measured value cannot be brought into its window | Decides whether the system is handed over conditionally or not at all |
Procedure
Confirm the pre-start checklist is complete and initialed before you close a disconnect. Read it rather than glancing at it: line-set decision, drain trap and safety switch, breaker against nameplate MOCP, low-voltage against this equipment's diagram. Acceptance: every line initialed by the installer. Stop rule: an incomplete sheet stops commissioning here and goes back to the lead installer. Wrong looks like energizing to "see what it does," which on a miswired reversing valve or a swapped compressor lead costs the compressor in the first minute.
Measure indoor airflow before you form any opinion about charge. Read total external static pressure across the air handler with a manometer at the manufacturer's stated port locations, then convert to airflow using that unit's own blower table at the installed tap or speed setting. Acceptance: total external static at or below the manufacturer's rated maximum, and airflow between 350 and 450 cfm per ton with 400 as the default, sitting deliberately near 350 only on a humid-climate setup where latent removal is the priority. Stop rule: outside that band, correct the airflow and do not touch the charge. Wrong looks like a delta T that "seems fine" standing in for a measurement.
Verify supply voltage, rotation and current before you judge anything thermal. Line voltage at the contactor within the nameplate range, compressor amps against nameplate RLA once running, and on any three-phase scroll confirm rotation immediately, because a scroll run backwards makes a distinctive noise, does not pump, and is damaged within minutes. Acceptance: voltage in range, amps at or below RLA, rotation correct. Stop rule: reversed rotation means kill it at once and swap two line leads before restarting; do not let it run to "confirm."
Weigh the charge to the nameplate figure plus this manufacturer's line-length adder. Measure the actual line-set length rather than using the sold estimate, and take the adder in ounces per foot from this manufacturer's instructions for the installed liquid line size. Acceptance: a scale reading recorded, not an estimate. Wrong looks like carrying another brand's adder across, which is the most common way a system arrives at startup heavy. Weight is the starting point; the governing reading in step 6 finishes the job.
Let the system stabilize before you read anything, and know which reading governs. Run at least 15 minutes at steady state with the space near setpoint. On a TXV or EEV system subcooling governs and superheat is a sanity check; on a fixed-orifice system superheat governs and must be evaluated against the manufacturer's chart for the outdoor dry bulb and the indoor wet bulb. Acceptance: readings stable across two consecutive checks five minutes apart. Wrong looks like reading at minute three, when suction is still falling and every number is going to move.
Trim to the governing reading in small increments, recording the reason for each move. Measure liquid pressure and liquid line temperature at the service valve, convert pressure to saturation temperature from the refrigerant's pressure-temperature chart, and subtract. Acceptance: subcooling inside the nameplate window, typically stated as a value plus or minus a tolerance. Wrong looks like adding or recovering half a pound at a time and chasing the reading past the window in both directions. Any refrigerant removed goes into a recovery cylinder.
Check the temperature split against the entering wet bulb, not against a fixed number. Measure return and supply dry bulb close to the coil, and the return wet bulb. Acceptance: a split consistent with the manufacturer's or a published sensible-split table for that entering wet bulb. The familiar figure near a 20 F drop is a rule of thumb at roughly 400 cfm per ton and moderate indoor humidity; at a high entering wet bulb more of the capacity goes to latent removal and the same correct system produces a smaller sensible split. Wrong looks like adding charge to force a 20 F split on a humid day.
Prove the safeties and the sequence, then hand over. Cycle the thermostat through each stage, confirm compressor lockout timing, flood the condensate pan to prove the safety switch opens, and on a heat pump force a defrost cycle and confirm the reversing valve shifts and auxiliary heat locks out as configured. Acceptance: each function observed, with the observed behavior written down. Wrong looks like assuming a factory-set control is correct. Do not flood a pan above a finished ceiling without a bucket staged under the trap.
The record this produces
A startup sheet carrying values, not check marks, filed against the equipment record:
- Total external static pressure and the blower table airflow it converts to, plus cfm per ton
- Supply voltage, compressor amps, nameplate RLA
- Measured line-set length, adder used, total charge weighed in, any amount recovered during trim
- Liquid pressure, saturation temperature, liquid line temperature, computed subcooling, and the nameplate target
- Suction pressure, saturation, suction line temperature, computed superheat
- Return dry bulb, return wet bulb, supply dry bulb, computed split
- Defrost and safety observations, and the tech's name and time
The warranty claim department reads this sheet first when a compressor fails inside its term. The next tech reads it as the baseline that says whether today's numbers have moved. The manager reads static pressures across the book to find the crews that are installing into ductwork nobody sized.
Worked pass: 3-ton R-410A split with a TXV
Step 1 clean. Step 2 FAILS. Total external static measures 0.92 in wc against a rated maximum of 0.50 in wc for this air handler. At 0.92 in wc the blower table gives roughly 950 cfm, which on 3 tons is 950 / 3 = 317 cfm per ton, below the 350 floor. Stop rule taken: charge untouched.
Cause found in two parts. A high-MERV filter had been dropped into a rack sized for the old unit's face velocity, and one flex return was crushed behind a joist. Filter moved to a properly sized media cabinet and the return re-supported. Static re-measured at 0.46 in wc, blower table gives about 1,180 cfm, so 1,180 / 3 = 393 cfm per ton. Inside the band, so step 2 now passes and commissioning continues.
Step 4: line set measures 40 feet against a nameplate charge of 7 lb 0 oz good for 15 feet. This manufacturer's adder for the installed 3/8 inch liquid line is 0.6 oz per foot beyond 15 feet, so 25 ft x 0.6 = 15 oz. Total weighed in: 7 lb 15 oz.
Step 6, after 15 minutes stable: liquid pressure 365 psig, which on the R-410A pressure-temperature chart is about 110 F saturated. Liquid line temperature 95 F. Subcooling is 110 - 95 = 15 F against a nameplate target of 10 F plus or minus 2. Above the window, so refrigerant is recovered in roughly 4 ounce increments with a 10 minute stabilization between each. After two increments the liquid line reads 99 F at the same 365 psig, so subcooling is 110 - 99 = 11 F. Inside the window. The 8 ounces recovered are logged to a cylinder number.
Step 7: return 75 F dry bulb, 63 F wet bulb, supply 56 F dry bulb, a 19 F sensible split. Consistent with that entering wet bulb at 393 cfm per ton. Had the return been at 67 F wet bulb, a smaller split would have been the correct result and the charge would not have been the reason.
Checking the pass against the sections above: step 2's band is 350 to 450 cfm per ton and both the failing figure (317) and the passing figure (393) are stated against it, in the same unit, computed from the same 3 tons. The subcooling target used in step 6 is the nameplate window named in that step, not a new number introduced by the example, and the recovered refrigerant went to a cylinder as the safety section requires rather than being vented during the trim.
References
- Manufacturer's installation and service instructions for nameplate charge, line-length adder, blower table, rated maximum external static and target subcooling
- ACCA Standard 5, HVAC quality installation specification, for airflow and charge verification practice
- 40 CFR Part 82 Subpart F for the venting prohibition that governs refrigerant removed during a charge trim
- See related: the system changeout day-of-install standard, and the thermostat replacement and configuration handover SOP