What an older built-in costs to run, and when that matters

Where the electricity actually goes

The compressor is most of it, but not all of it, and the remainder is larger than people expect. Sub-Zero built-ins use dual refrigeration, which means two entirely separate sealed systems, one for the fresh food compartment and one for the freezer, with no air passing between them. Two compressors sounds like twice the bill and is not: each is sized for its own compartment and each runs only when that compartment asks for cooling. The condenser fan runs alongside whichever compressor is working. The evaporator fans are small but turn nearly all the time. Then come the heaters, which are the part nobody counts, because a refrigerator contains several of them: a defrost heater that clears the coil on a schedule, a drain heater on models that carry one, the mold heater firing on every ice harvest, anti-condensation heat around door edges and mullions on some models, and on a wine or glass-door cabinet a pane heater working to keep the glass clear. Anything warm inside the box is paid for twice, once to power it and once to pull the heat back out, which is why a cabinet still lit by incandescent lamps costs a little more than the bulb wattage suggests.

• Compressors: the largest single share, and on dual refrigeration there are two, each running independently
• Condenser fan, turning whenever a compressor is running
• Evaporator fans: small draw, but close to continuous
• Defrost heater: a heavy load for a short time, repeated on a schedule
• Ice maker mold heater on every harvest, plus a fill tube heater where one is fitted
• Anti-condensation heat at door edges and mullions, and glass heaters on wine and glass-door cabinets
• Heat generated inside the cabinet, which the sealed system then has to remove again

The number worth measuring is run time, not kilowatt-hours

Metering a built-in is awkward on purpose. It normally sits on a dedicated circuit whose outlet is behind or above the cabinet, sometimes hardwired, and a clamp meter on the supply is a technician's tool and a technician's access. You do not need either. Sit in the kitchen for an hour with a phone and note when the compressor starts and when it stops. What you end up with is a duty cycle, and duty cycle is the whole story, because a sealed system either satisfies its compartment with margin left over or it does not. The power of the measurement is that you are comparing the appliance against itself rather than against a benchmark, and there is no single correct percentage that fits every kitchen: the same unit legitimately runs longer on a hot inland afternoon in September than it does in February. But a machine that used to shut off and now never does has changed, and something caused that change. Two practical notes. On dual refrigeration, work out which compartment you are listening to before drawing conclusions, since the two compressors run to their own schedules. And when you convert run time into money, use the highest tier or on-peak rate printed on your own bill rather than an average, because every extra hour of compressor time is billed at the margin, not at the mean.

• Time the compressor across an hour, twice: once on a cool morning and once on a warm afternoon
• Compare the unit to its own past behavior, not to a figure from an article
• Identify which compartment's compressor you are hearing before deciding anything
• For dollars, use the top tier or on-peak rate from your own statement
• Skip plug-in meters unless the outlet is genuinely reachable and rated for the load, which on most built-ins it is not
• Write the result down with the date, because the value of this reading is entirely in the comparison you make a year from now

Four things that buy run time back, none of them a new appliance

A matted condenser is the largest reversible cause of extra run hours on this equipment, because heat that cannot leave through the coil leaves through longer compressor cycles instead. A door seal that has lost magnet pull is next, and it costs twice: warm humid air enters continuously, and the moisture it carries loads the evaporator with frost that the defrost circuit then burns more energy clearing. Third is the installation, which belongs on an energy list and not only on a fault list. A grille that cannot draw the airflow the model was rated for does not fail in any way you can point at; it just makes the compressor work longer for the same result, every warm afternoon, for as long as the appliance sits in that opening. Heat arriving from next door does the same thing, whether that is an oven, a dishwasher or afternoon sun landing on the cabinet run. Fourth is the setpoint, and the principle is short even though it deserves its own article: every degree colder than the food actually requires is bought with compressor hours, every day, forever. Add one more that is not on anybody's list, which is frost that returns between defrost cycles. A furred evaporator is an insulated evaporator, and an insulated evaporator makes the compressor run to no purpose at all.

• Condenser condition: hold a flashlight to the grille, and if the fins look like felt rather than a comb, that is run time you are paying for
• Gasket pull measured on all four sides, and door squareness checked rather than eyeballed
• True clearance at the grille, measured against what the model requires rather than judged by how the cabinetry looks
• Neighboring heat: an oven, a dishwasher or afternoon sun on the cabinet run
• Frost returning between defrost cycles, which points at the heater, the thermistor or the seal
• A setpoint colder than the contents need, held three hundred and sixty-five days a year

When the number actually matters

Three situations, and only three. The first is diagnostic, and it is the one worth caring about: rising run time is the earliest warning any refrigerator gives, and it usually appears weeks before a temperature complaint. Out here the calendar makes that concrete. Marginal units get through an entire summer and then fail in a block during the first Santa Ana of the fall, when hot dry air moves through the house for days at a stretch, and the run-time change that predicted it was visible in August. The second is the appliance nobody opens. A wine cabinet in a room people walk past, or a second refrigerator holding four items, is running around the clock for very little, and the arithmetic on that is genuinely different from the kitchen unit. The third is the repair-or-replace conversation, where consumption belongs as one line and not as the deciding one, because replacing a fitted appliance means matching a cabinet opening and, in the newer kitchens up here, a custom panel front. One honest caveat closes the subject: if an older unit draws what it draws because of the year it was designed, and everything about it is otherwise healthy, no repair changes that. Efficiency is built in at the factory. What service can recover is the difference between the appliance working properly and the appliance working hard, and on a neglected unit that difference is bigger than the difference between model years.

• Rising run time as an early warning, checked before the hot weeks rather than during them
• Units in rooms nobody enters, doing very little work for a full-time draw
• Consumption as one input to a replacement decision, never the whole of it
• The gap between a healthy old unit and a neglected one, which is often larger than the gap between generations