Cold Weather Performance Complaint Visit

Purpose

This standing instruction guarantees that a heat pump accused of not keeping up on the first hard cold night is measured against its own published capacity at that outdoor temperature before any part is quoted.

Most of these visits end with equipment that is performing exactly as published and a setting, a load or a duct system that is not. A shop with no way to measure delivered capacity has only two moves on that call, guess or replace, and both of them bill a customer for a working machine. The measurement itself is cheap: airflow off the blower table at a measured static, a supply and return pair taken with one instrument, and one multiplication. What makes it worth doing is that it produces a number the customer, the next tech and the manufacturer can all argue with.

Safety actions that gate the work

• This procedure disables the backup heat for the measurement, which leaves the house on compressor heat alone. Do not run it on a house already below a safe indoor temperature, or where an infant, an elderly occupant or anyone medically vulnerable is present. Where either applies, the measurement waits for a warmer day or is done with the occupants elsewhere.
• Open and lock the strip-heat breakers and prove them dead before your hands are in the air handler, proved on a known live source before and after per NFPA 70E-2021, 120.5, with work practices at 29 CFR 1910.333(b)(2), written for qualified persons under 1910.332. Elements stay hot after the circuit opens.
• Write the backup heat down as disabled the moment you disable it, with the restore step named. A visit that ends with strips still locked out on a January night is a far worse outcome than the complaint you came for.
• Static pressure taps go in with the blower off. A probe pushed into a running return can be pulled into the wheel, and a hole drilled in an energized cabinet finds the wiring you did not know was behind it.

Scope

Covers the measurement visit on a residential or light-commercial air-source heat pump whose owner reports that it is not keeping up in cold weather, from complaint intake through a written finding of equipment, setting, load or distribution.

Does not cover the corrective work the finding points to: lockout and backup sizing belong to the backup heat SOP, charge to the charge verification SOP, defrost to the defrost control SOP, and duct sizing to the duct leakage and design documents. Does not cover a system that is not running at all, which is a no-heat call. Does not cover the customer conversation about what normal looks like, which the heat pump handover SOP owns.

Roles and handoffs

Procedure

1. Separate failing to hold from failing to recover before you measure anything. Acceptance: the setpoint, what the house is actually holding, over how many hours, and whether it climbs back during the day. Wrong looks like measuring first and discovering afterwards that the house recovers every afternoon, which points at load and setback rather than capacity. Hazard: none at this step, it is a conversation, and it is the step that decides whether the rest of the visit is worth running today.

2. Disable the backup heat and record that you have. Acceptance: the strip breakers open and proved dead, or the control's auxiliary disabled, plus a written line naming step 7 as the restore. Wrong looks like measuring supply air with strips running, which measures the strips and reports a heat pump that is performing beautifully. Stop rule: the occupant gate in the safety block is checked before this step, not after. Hazard: line-voltage strip terminals sit beside the low-voltage board, and the elements stay hot after the breaker opens, so prove dead and let them cool before reaching past them.

3. Measure total external static across the blower and read airflow off the blower table. Acceptance: a static in inches of water column measured at the equipment with the blower off during probe placement, a value inside the published range of that model's blower table, and a CFM read from the row for that static and speed tap. Wrong looks like a static above the table's top published row, where no airflow can be read and every capacity figure after it is invented. Stop rule: static outside the table stops the measurement and the airflow restriction gets fixed first. Hazard: drilling and probing at the cabinet, so the blower is off and the disconnect open while probes go in.

4. Take supply and return dry bulb at the equipment with ONE instrument moved between the two points. Acceptance: both readings and the difference, with the instrument named. A fixed offset in one instrument cancels in a difference, which is why one instrument beats two; two instruments contribute independent spreads that combine in quadrature and make the rise less certain, not more. Wrong looks like reading supply at a distant register, which includes duct loss the equipment did not cause. Hazard: probing at the supply plenum with the blower running, so keep fingers and lanyards out of the cabinet opening.

5. Compute delivered sensible capacity. Multiply 1.08 by the CFM by the temperature rise. The 1.08 is derived for standard air at about 0.075 lb per cubic foot near sea level; air thins with elevation, so at roughly 5,000 ft the constant is nearer 0.90 and using 1.08 there overstates capacity by about a fifth. Acceptance: a Btu/h figure with the constant used and the site elevation stated beside it. Wrong looks like carrying 1.08 to a mountain town. Hazard: none, this is arithmetic done at the truck.

6. Compare against the extended capacity table at THIS outdoor temperature and entering air, and decide with the measurement's own uncertainty in hand. Acceptance: the table value read, the measured value as a percentage of it, and a verdict against a 15 percent band. That band is chosen because the airflow term is roughly 10 percent of reading when CFM comes from a blower table, and the rise term is the instrument's repeatability, commonly about 0.5 F, as a percentage of the rise you actually measured; independent spreads combine in quadrature, so the two together land near 10 percent and a 15 percent gate sits outside them. Wrong looks like comparing against the nameplate rating, which is published at one mild outdoor condition. Hazard: none, desk step.

7. Restore the backup heat and prove it before you write anything. Acceptance: strip breakers closed, one auxiliary stage watched energizing at the sequencer, the lockout value read back and confirmed unchanged from what you found, and a written finding naming equipment, setting, load or distribution. Wrong looks like closing the breaker and assuming; a sequencer that did not pick up leaves the house worse than you found it. Stop rule: any auxiliary stage that does not energize keeps you on site. Hazard: this is the step that puts the highest-current circuit in the house back with the family present, so stand to the hinge side when you close it, confirm the blower runs before elements energize, and watch the stage drop out at the end of the call.

The record this produces

A capacity worksheet filed to the equipment:

• Setpoint, actual indoor temperature, hours held, and whether it recovers in the day
• Backup heat disabled, with the time, and restored, with the time
• Total external static, whether it fell inside the blower table, and the CFM read with its row and speed tap
• Supply, return and the rise, with the single instrument named
• The constant used, the site elevation, and the computed Btu/h
• The table value read, at what outdoor and entering air condition, and the measured percentage of it
• The auxiliary stage proved at the sequencer, and the lockout value as read back
• The finding, named as equipment, setting, load or distribution, and the SOP it refers to

The follow-on visit quotes against this sheet rather than against the complaint, which is the difference between selling a fix and selling a guess. The manager reads the finding column across a winter: a run of settings findings is a commissioning gap, and a run of distribution findings is a sales gap.

Worked pass: 18 F outdoor, house holding 66 F against a 70 F setpoint, single-stage heat pump with electric backup

Step 1: setpoint 70 F, house holding 66 F, has been that way since roughly midnight, and it does climb back to setpoint by mid-afternoon. So it holds a deficit rather than falling away, which points at capacity against load rather than at a failure.

Step 2: occupant gate checked, no vulnerable occupant and the house is at 66 F, so the measurement may proceed. Strip breakers opened and proved dead at 09:14, restore noted against step 7.

Step 3 FAILS. Total external static measures 0.92 in w.c. against a blower table whose top published row is 0.8 in w.c. No CFM can be read, so the stop rule fires and the airflow is fixed first. Filter found loaded and one return branch damper found closed. Filter replaced, damper opened, static re-measured at 0.55 in w.c., inside the table, and CFM read as 1,150 at that static on the installed speed tap.

Step 4: one digital thermometer moved between the two points. Return 68 F, supply 90 F, so the rise is 90 minus 68, which leaves 22 F.

Step 5: site is near sea level, so the constant is 1.08. Capacity is 1.08 times 1,150, which gives 1,242, times 22, which leaves 27,324 Btu/h.

Step 6: the extended capacity table for this model at 18 F outdoor and 68 F entering air reads 28,500 Btu/h. Measured over table is 27,324 divided by 28,500, which gives 0.959, so 95.9 percent. The uncertainty check: the airflow term is about 10 percent of reading, the rise term is 0.5 F on a measured 22 F rise, which is 2.3 percent, and combined in quadrature those give about 10.3 percent. A 4.1 percent shortfall sits well inside both the 15 percent gate and the measurement's own uncertainty, so the equipment is performing.

Step 7: strip breakers closed at 10:41, one auxiliary stage watched energizing at the sequencer, lockout read back at 15 F, unchanged from as-found. Finding written as a setting problem: the balance point worksheet filed to this equipment gives a load line of 700 Btu/h per degree F and a balance point of 21 F, both read from that sheet. At 18 F the load is 700 times 65 minus 18, which leaves 47 F, giving 32,900 Btu/h, against 27,324 Btu/h delivered, so the house is short 32,900 minus 27,324, which leaves 5,576 Btu/h, and the auxiliary that would cover it is locked out at 15 F, 6 F below the balance point. Referred to the backup heat sizing and lockout SOP, which owns the correction.

Checking the pass against the sections above: step 6's acceptance names a table value, a percentage and a verdict against 15 percent, and the run prints 28,500, the division, 95.9 percent and the uncertainty terms rather than asserting the verdict. The 1.08 in step 5 is used at a near-sea-level site, which is the condition step 5 states it was derived under. The load line and balance point in step 7 are read from the equipment's existing worksheet and labelled as read, not derived here.

References

• Equipment manufacturer's blower performance table for airflow at a measured external static, and the extended capacity table for capacity at a given outdoor and entering air temperature
• ACCA Manual J for the load line and design condition that a balance point worksheet is built from
• 29 CFR 1910.333(b)(2) and NFPA 70E-2021, 120.5, for de-energizing and proving dead at the strip circuits
• See related: backup heat sizing and lockout configuration, charge verification by weight and superheat, defrost control diagnosis on a heat pump, heat pump customer education at handover