Pre-cooling a house before peak hours saves money only when three things line up: a wide enough gap between peak and off-peak prices, a house tight enough to coast through the peak, and equipment that does the extra work efficiently. It rarely saves energy; it moves cooling into cheaper hours and usually costs a little extra energy, so the price gap has to pay for that penalty. A common starting point is 2 to 3°F cooler for two to three hours before the peak and 3°F warmer during it. In winter, pre-heat in 1°F steps with the heat pump alone, because a pre-heat that switches on resistance heat strips can cost more than it saves.
This guide puts that answer into numbers. It covers the evidence (utility guidance, California code modeling, a small field test and a PNNL winter study), a single ratio that tells you whether pre-conditioning can pay on your rate, a coast test for your own house, and worked summer and winter schedules on LG&E’s time-of-day rate in Kentucky, which has a summer afternoon peak and two winter peaks a day.
How pre-cooling and pre-heating work
Pre-cooling runs the air conditioner harder before the peak window so the house enters the peak cooler than usual, then lets the indoor temperature drift up, or “float,” while the system runs less. Pre-heating does the same in winter, warming the house a little before the peak and letting it drift down during it. Both use the house itself, its air, walls, floors and furniture, as a small thermal battery.
A paper by Kristin Heinemeier of Frontier Energy, from ACEEE’s 2022 Summer Study and posted on ACEEE’s website, defines pre-cooling this way: the home is cooled below its usual temperature in the hours just before the utility peak and then allowed to float through the peak with less, or no, cooling energy. It notes that a programmable thermostat can do this as well as a connected one.
Why it usually costs a little energy
A house gains or loses heat faster the larger the gap between indoor and outdoor temperature. A pre-cooled house is cooler than usual during the hottest hours, so heat leaks in faster; a pre-heated house loses heat faster. PSEG Long Island makes the winter version of this point on its time-of-day tips page (Wayback Machine capture, March 16, 2026): the warmer a home is, the faster thermal energy is lost to the outside.
That extra leakage is the energy penalty. Some of the coolness stored before the peak is lost before it is used, and the system may overshoot when it recovers afterward. The penalty is small in a tight, heavy house and large in a leaky, light one.
What decides how long a house can coast
- Envelope. Insulation, air sealing and windows set how fast heat moves in or out.
- Thermal mass. Masonry, tile floors, plaster and a full house of furniture store more heat per degree than a lightweight frame house.
- Solar gain. West-facing glass in a late-afternoon peak works against pre-cooling.
- Ducts. Ducts in a hot attic add heat to the air they carry, which lowers what the system delivers to the rooms.
- Peak length. A three-hour peak is far easier to coast through than a five-hour one.
What the evidence says
Independent evidence on pre-cooling savings is thinner than the advice suggests. Most published figures come from thermostat makers, and this guide leaves those out. What follows is the utility guidance, the modeling done for California’s building code, a small field test, and a PNNL winter study.
What utilities advise
Arizona Public Service, whose time-of-use plans have a 4 to 7 p.m. weekday peak, gives three steps on its pre-cooling page: set the thermostat “a few degrees lower than usual for several hours during the day,” raise it “several degrees” from 4 to 7 p.m., and set it as you like after 7 p.m. PSEG Long Island, whose capture lists a 3 to 7 p.m. weekday peak, is more specific: at noon, three hours before the peak, set the thermostat three degrees below your usual setting; at 3 p.m., three degrees above it; at 7 p.m., back to usual. It adds that results vary with comfort preference and the size of the home.
What California’s code modeling found
The ACEEE paper summarizes a 2020 analysis for a possible pre-cooling credit in California’s Title 24 building code. Modeling two prototype homes, it found the best setting in most climate zones was to pre-cool to 75°F from noon to 4 p.m. and let the house float to as high as 83°F from 4 to 9 p.m.
The results are a useful reality check. For the 2,100-square-foot home, pre-cooling raised total kWh by 1% to 2% in every climate zone listed. Under the PG&E time-of-use rate modeled, the bill changed by between a 1.7% increase and a 2.7% decrease. Only under a hypothetical rate with free off-peak power and $1.33 per kWh at peak did savings reach 8% to 51%. The paper’s conclusion: pre-cooling can help the grid’s load shape, but energy savings are not expected, and bill savings are not expected either under time-of-use rates that are not “peaky” enough.
What small field tests showed
The same paper reports a one-week test of researcher-controlled pre-cooling in two Southern California homes, pre-cooling to 72 to 76°F from noon to 4 p.m. and holding 80°F or higher from 4 to 9 p.m. Both homes shifted cooling ahead of the peak, but two findings matter for homeowners. Automatic pre-cooling on mild days may increase energy use unnecessarily, and higher pre-cooling setpoints used less energy overall. The thermostats’ own optimization routines also changed the test settings from day to day, which the researchers did not know at first.
A separate 2021 analysis, also summarized there, found that choosing a single well-picked pre-cooling setting for a climate captured nearly as much value as a perfect day-by-day forecast. A simple, consistent schedule loses little.
What a winter heat pump study found
PNNL tested demand flexibility with ductless heat pumps in three homes in Cordova, Alaska, and presented the results to DOE’s Better Buildings network in November 2024 (slides). The events lowered the setpoint by 2°F or 6°F for one or two hours, some with one or two hours of pre-heating at 2°F above the normal setpoint.
Hourly energy use fell 6% to 35% during the test hours, and pre-heating increased the peak reduction in every test. Larger setbacks were tolerated as long as events lasted no more than two hours; longer events made occupants more likely to override. That is a useful limit for homes facing a four-hour winter peak.
What certified smart thermostats must do
ENERGY STAR’s smart thermostat criteria require field-verified run-time reductions of at least 8% for heating and 10% for cooling (lower 95% confidence limits of weighted national averages), the ability to work with utility programs, and a way for the consumer to override grid requests. For heat pump installations, the thermostat must report electric resistance heat use. None of these criteria measure pre-cooling or pre-heating savings on a time-of-use rate.
When the price gap pays for the extra energy
Pre-conditioning pays when the money saved on the kWh moved out of the peak is larger than the cost of the extra kWh it takes. That gives one formula and one ratio you can work out from your tariff sheet.
Net saving = kWh moved out of the peak x (peak price minus off-peak price) minus extra kWh x off-peak price.
Rearranged, the break-even point is the price gap divided by the off-peak price. That ratio is the most extra energy you can spend, per kWh moved out of the peak, before pre-conditioning stops paying.
| Rate (season) | Pre-peak price per kWh | Peak price per kWh | Gap divided by pre-peak price | What it means |
|---|---|---|---|---|
| LG&E RTOD-Energy (all year) | $0.09312 | $0.19307 | 1.07 | Each kWh moved can carry about 1 extra kWh before savings vanish |
| Culni example rate, fictional (pre-cooling in day hours) | $0.14 | $0.31 | 1.21 | Similar room for error |
| SCE TOU-D-4-9 (summer weekdays, 2024) | $0.28 | $0.51 | 0.82 | Moderate room |
| PG&E TOU-C (summer, 2024) | $0.44 | $0.52 | 0.18 | Only a small penalty is affordable |
| PG&E TOU-C (winter, 2024) | $0.39 | $0.42 | 0.08 | Almost any penalty erases the saving |
The table explains the California modeling result. On a rate like PG&E’s TOU-C as listed for 2024, the peak costs only 8% to 18% more than other hours, so a 1% to 2% rise in total kWh can eat the whole saving. On LG&E’s time-of-day rate, the peak price is roughly twice the off-peak price, so the same penalty barely dents it.
Example (fictional). A house on LG&E RTOD-Energy on a hot summer weekday. Normally the air conditioner holds 75°F and uses 3.0 kWh from 11 a.m. to 2 p.m., 6.0 kWh in the 2 to 6 p.m. clock-time peak and 3.0 kWh from 6 to 9 p.m.: 12.0 kWh.
- Tight house, pre-cooled to 72°F by noon, floating to 78°F in the peak: 6.5 kWh from 11 a.m. to 2 p.m., 3.0 kWh in the peak, 3.8 kWh from 6 to 9 p.m. (recovering to 75°F). Total 13.3 kWh, so the penalty is 1.3 kWh.
- Net saving: 3.0 kWh moved out of the peak x $0.09995 = $0.30, minus 1.3 kWh x $0.09312 = $0.12, leaves about $0.18 for the day.
- Leaky house, same settings: it warms so fast that the air conditioner restarts at 78°F partway through the peak. Only 2.0 kWh leaves the peak and the penalty is 2.5 kWh. Net: 2.0 x $0.09995 = $0.20 minus 2.5 x $0.09312 = $0.23, a loss of about $0.03.
Over 20 hot weekdays, the tight house saves about $3.60 before riders and taxes; the leaky house loses a little. These are energy charges only.
The dollar figures are small because this is one appliance on a moderate gap. They grow with a larger air conditioner, a hotter climate or a wider price gap, and they shrink, or turn negative, with a leaky house or a mild day.
How many degrees and how early: test your house first
The right depth depends on how fast your house warms with the cooling off, and you can measure that yourself in one afternoon. Two numbers decide the setting: the drift rate (degrees per hour with the system off) and the comfort range you will accept during the peak.
The coast test
- Pick a hot weekday that is typical of your summer peak days.
- At the start of the peak window, raise the setpoint far enough that the air conditioner stays off, and note the indoor temperature.
- Write down the indoor temperature every hour until the window ends or the house reaches the highest temperature you would accept.
- Divide the total rise by the hours elapsed. That is your drift rate for that weather.
- Repeat on a hotter day if you can, because drift speeds up with outdoor temperature and afternoon sun.
Turning the drift rate into a setting
Pre-cool by enough that the house reaches your comfort ceiling at about the end of the peak. The arithmetic: pre-cool depth = drift rate x peak hours minus the float you accept above your normal setpoint.
| Drift rate with cooling off | 3-hour peak, 3°F float accepted | 4-hour peak, 3°F float accepted | 4-hour peak, 4°F float accepted |
|---|---|---|---|
| 1°F per hour | 0°F (no pre-cool needed) | 1°F | 0°F |
| 1.5°F per hour | 1.5°F | 3°F | 2°F |
| 2°F per hour | 3°F | 5°F (too deep; plan on partial cooling) | 4°F |
| 3°F per hour | 6°F (too deep; plan on partial cooling) | 9°F (coasting not practical) | 8°F (coasting not practical) |
These are starting points from simple arithmetic, not research findings. Drift is not perfectly steady: it slows as the house’s mass warms and speeds up when the sun reaches west-facing windows.
When the required depth goes beyond about 3 to 4°F, do not chase it. Pre-cool by 2 to 3°F, raise the peak setpoint by 3 to 4°F, and accept that the system will run for part of the peak. That still moves some kWh, and it avoids the energy penalty of a deep pre-cool that leaks away.
Comfort and humidity during the float
Temperature is not the only thing that changes during a float. An air conditioner removes moisture only while it runs, so a house coasting with the system off can feel stickier at 78°F than the number suggests, especially in a humid climate. If the float feels worse than expected, lower the ceiling by a degree and run a ceiling fan, which makes a room feel cooler without changing its temperature. In the ACEEE paper’s own example home, ceiling fans were the first response before the air conditioner went back on.
How early to start
Start early enough for the system to pull the house down by the full depth, and no earlier. Time how long your system takes to lower the house 2 or 3 degrees on a hot afternoon; that, plus a margin, sets the start. PSEG’s example starts three hours ahead; the California modeling used four. Starting too early lengthens the period when the house is cooler than usual, and that is when the penalty builds.
A summer schedule for an afternoon peak
LG&E’s tariff sets its rating periods in Eastern Standard Time all year (Sheet No. 6.1). From April through October, the weekday on-peak period is 1 to 5 p.m. EST, which the LG&E and KU time-of-day page shows as 2 to 6 p.m. on daylight-saving clocks. Weekends are off-peak all day. On RTOD-Energy (Sheet No. 6, effective September 1, 2026), on-peak energy costs $0.19307 per kWh and off-peak $0.09312, base charges before riders and taxes.
The utility page also states the risk plainly: “If you do not shift the time of day that you use energy or you consume energy during on-peak times, your bill may go up.” Whether this rate suits a household at all depends on how much of its use can move, which is a separate decision from the thermostat schedule.
| Clock time (summer weekday) | Thermostat setting, from a 75°F normal | Why |
|---|---|---|
| Until 11 a.m. | 75°F | Normal |
| 11 a.m. to 2 p.m. | 72 to 73°F | Pre-cool in off-peak hours |
| 2 to 6 p.m. | 78°F | Peak; the house floats |
| 6 to 7 p.m. | 76°F, then 75°F | Recover in two steps to limit the post-peak surge |
| Weekends (a weekday holiday counts as a weekday) | 75°F | All off-peak on this rate |
A five-hour evening peak, worked through
On a rate with a 4 to 9 p.m. peak, such as the fictional example rate on the example home page, the peak is five hours long and runs into the evening, when people are home and the sun is low. Coasting all five hours is unrealistic in most houses. A practical version pre-cools from about 1 to 4 p.m., floats to the comfort ceiling, and lets the system hold that ceiling for the last part of the peak.
Example (fictional). The example rate is fictional: 8 cents per kWh off-peak (9 p.m. to 7 a.m.), 14 cents in day hours (7 a.m. to 4 p.m.) and 31 cents in the weekday peak (4 to 9 p.m.). Normally the air conditioner holds 75°F and uses 4.5 kWh from 1 to 4 p.m., 8.0 kWh from 4 to 9 p.m. and 1.5 kWh from 9 to 11 p.m.
- Normal cost: 4.5 x $0.14 = $0.63, plus 8.0 x $0.31 = $2.48, plus 1.5 x $0.08 = $0.12, for $3.23.
- Pre-cooled to 72°F from 1 p.m., floating to 78°F and held there from about 7 p.m.: 7.0 kWh from 1 to 4 p.m., 4.0 kWh in the peak and 3.4 kWh from 9 to 11 p.m. while the house returns to 75°F. That is 14.4 kWh against 14.0, a penalty of 0.4 kWh.
- Pre-cooled cost: 7.0 x $0.14 = $0.98, plus 4.0 x $0.31 = $1.24, plus 3.4 x $0.08 = $0.27, for $2.49. The day costs about $0.74 less.
The saving is larger than in the LG&E example because the peak price is more than double the day price, and because the system still runs at the end of the peak instead of trying to coast through all five hours.
Winter pre-heating for morning and evening peaks
Winter pre-heating can pay on a rate with winter peaks, but with a heat pump it has a trap: a pre-heat that raises the setpoint too fast switches on the resistance heat strips, and strip heat costs so much more per unit of heat that it can wipe out the saving. The rule is small steps, heat pump only, and an early start.
The winter windows
From November through March, LG&E’s weekday on-peak periods are 6 to 10 a.m. and 6 to 10 p.m. EST, with off-peak from 10 a.m. to 6 p.m. and 10 p.m. to 6 a.m. (Sheet No. 6.1). The morning peak is the hard one: it starts in the coldest part of the day, when a heat pump has the least capacity, and it catches any recovery from an overnight setback.
Why heat strips change the math
Heat strips are electric resistance elements in the air handler. Santee Cooper’s Heat Strips 101 (January 13, 2026) says they switch on “when you raise your thermostat several degrees at once” and commonly when outdoor temperatures are anywhere from 32°F to 40°F, and it advises raising the temperature by no more than one or two degrees at a time. Alabama Power’s explainer says strip heat engages when the indoor temperature is outside a 1.5 to 2 degree range of the setting, and that strips use roughly three to four times the energy a heat pump uses for the same heat (Alabama Power, December 16, 2024). Google’s Nest help page puts auxiliary heat at about 2 to 5 times the cost of running the heat pump.
Strips turn each kWh into one kWh of heat. A heat pump moves more heat than it uses; its coefficient of performance (COP) is the ratio. ENERGY STAR’s cold-climate criteria require a COP of at least 1.75 at 5°F, and the COP is higher in milder weather. Dividing the price by the COP gives the cost of each kWh of heat:
| Heat source | Off-peak ($0.09312 per kWh) | On-peak ($0.19307 per kWh) |
|---|---|---|
| Heat pump at COP 3.0 | 3.1 cents per kWh of heat | 6.4 cents |
| Heat pump at COP 2.0 | 4.7 cents | 9.7 cents |
| Heat pump at COP 1.75 (ENERGY STAR cold-climate minimum at 5°F) | 5.3 cents | 11.0 cents |
| Heat strips (one kWh of heat per kWh) | 9.3 cents | 19.3 cents |
The key comparison is the bottom-left cell against the right-hand column. Strip heat bought off-peak at 9.3 cents costs about the same as heat pump heat bought on-peak at a COP of 2.07 (19.307 / 9.312), and more than heat pump heat on-peak at any higher COP. Whenever the heat pump is running at a COP above about 2, a pre-heat that brings on the strips costs more than simply letting the heat pump run through the peak.
A winter morning schedule
| Clock time (winter weekday, standard time) | Thermostat setting, from a 68°F normal | Why |
|---|---|---|
| 10 p.m. to 4 a.m. | 68°F, or a setback of no more than 1 to 2°F | Avoids a large recovery later |
| 4 to 5 a.m. | 69°F | First pre-heat step, heat pump only |
| 5 to 6 a.m. | 70°F | Second step |
| 6 to 10 a.m. | 66 to 67°F | Morning peak; the house drifts down |
| 10 to 11 a.m. | 67°F, then 68°F at 11 a.m. | Recover in steps, off-peak |
| 3 to 6 p.m. | 69 to 70°F in two steps | Pre-heat in the warmest part of the day |
| 6 to 10 p.m. | 66 to 67°F | Evening peak |
The afternoon pre-heat is the easier of the two. Afternoons are usually the mildest part of a winter day, so the heat pump runs at its best COP, and 3 to 6 p.m. is off-peak on this rate.
Example (fictional). An all-electric house with a heat pump and 10 kW of heat strips, on LG&E RTOD-Energy, on a 25°F weekday morning. Normally it holds 68°F and the heat pump uses 4.0 kWh from 4 to 6 a.m., 8.0 kWh from 6 to 10 a.m. and 3.0 kWh from 10 a.m. to noon: 15.0 kWh.
- Stepped pre-heat, heat pump only: 6.5 kWh from 4 to 6 a.m., 4.0 kWh in the peak, 5.0 kWh from 10 a.m. to noon. Total 15.5 kWh, a penalty of 0.5 kWh. Net saving: 4.0 kWh moved x $0.09995 = $0.40, minus 0.5 x $0.09312 = $0.05, leaves about $0.35.
- Same pre-heat in one 2°F jump at 4 a.m.: the strips run for 30 minutes, 10 kW x 0.5 hours = 5.0 kWh, to deliver heat the heat pump would have delivered with about 2.0 kWh at an assumed COP of 2.5. The extra 3.0 kWh off-peak costs 3.0 x $0.09312 = $0.28, leaving about $0.07 of the saving.
- Overnight setback to 62°F, recovered at 6 a.m.: if the strips run 30 minutes in the peak, that alone is 5.0 kWh x $0.19307 = $0.97, more than the stepped pre-heat saves.
All figures are energy charges only, before riders and taxes.
Transition weeks, weekends and demand rates
LG&E’s page shows the winter peaks one hour later on the clock, 7 to 11 a.m. and 7 to 11 p.m., from the second Sunday in March through March 31 and from November 1 to the first Saturday in November, when clocks are on daylight time but the winter windows still apply. A thermostat schedule written in clock time needs moving for those weeks. Weekends are off-peak all day, so the pre-heat program should run on weekdays only, and a legal holiday on a weekday still counts as a weekday.
On LG&E’s other time-of-day rate, RTOD-Demand (Sheet No. 7, effective September 1, 2026), energy costs $0.06427 per kWh but the bill adds $4.32 per kW for the highest 15-minute demand of the month in any hour and $9.46 per kW for the highest in peak hours (Sheet No. 7.1), base charges before riders and taxes. There, a strip-assisted pre-heat at 5 a.m. still costs money even though it is off-peak: 10 kW of strips plus a 3 kW heat pump running together set a 13 kW base demand, 13 x $4.32 = $56.16 for the month, against 3 x $4.32 = $12.96 for the heat pump alone (hypothetical loads). On a demand rate, step the pre-heat even more gently.
Overnight setbacks and heat pumps
Generic setback advice does not account for heat strips. PSEG Long Island’s page, for example, suggests 58 to 62°F at night or when away and raising the heat about an hour before you get up. With a heat pump on a rate with a morning peak, that recovery lands at the worst time: the coldest hour, the start of the peak, and a large gap that invites the strips. Keep setbacks small, or end them before the morning pre-heat begins. Why strips come on, and how to check whether yours run too much, is covered in the guide to heat pump auxiliary heat.
When pre-conditioning backfires
Pre-conditioning loses money or comfort in predictable situations. Skip it, or make it shallower, when any of these apply:
- The price gap is small. If your gap-to-off-peak ratio from the table above is below about 0.2, a typical energy penalty takes most of the saving.
- The house drifts fast. A drift rate above about 2.5°F per hour on a four-hour peak means the system restarts partway through, as in the leaky house example.
- The day is mild. On a mild day the house would have needed little cooling in the peak anyway, so a pre-cool mainly adds energy; the California field test found the same.
- The peak is long and runs into the evening. Five-hour peaks ending at 9 p.m. leave too little stored coolness by the end.
- Heat strips are reachable. Any winter pre-heat step large enough to call for auxiliary heat.
- The rate has a demand charge. On a rate that bills the highest 15-minute kW, a fast pre-cool or a post-peak recovery that overlaps EV charging or a water heater element can set a new monthly peak. LG&E’s RTOD-Demand rate is an example, described in the winter section above.
- Someone at home needs a steady temperature. Infants, older adults and people with some medical conditions may not tolerate a 3 to 4°F swing. Comfort and health come first.
Thermostat settings that make it work
As the ACEEE paper notes, a plain programmable thermostat with a weekday schedule can run pre-cooling, and the same goes for pre-heating. A few settings decide whether it works as intended.
Schedule, not optimization, at first
The California field test found that the thermostats’ own optimization features quietly changed the researchers’ schedules. If your thermostat has features that adjust setpoints on its own, check whether they override a manual schedule, and turn them off for a test week so you can see what your schedule does by itself.
Heat pump settings for winter
Some thermostats that control heat pumps let an installer set an auxiliary heat lockout, the outdoor temperature above which strips are not allowed to run, and some offer automatic modes that trade comfort against strip use. Google’s Nest help page describes Heat Pump Balance with Max Comfort (the default), Balanced and Max Savings settings, and a manual lockout temperature when the feature is off. Other makers use different names. A licensed HVAC technician should set lockout temperatures, because they depend on the heat pump’s capacity at low temperatures.
Program events take priority
If your thermostat is enrolled in a utility demand response or peak time rebate program, the program’s events will override your schedule on event days, and an event can stack with your pre-cool or replace it. How those programs pay and what they control is covered in the guide to thermostat demand response programs.
Coordinating the thermostat with the rest of the house
In an all-electric home, the thermostat is one of several loads that move around the same windows. A heat pump water heater loads up before the same peaks, an EV may charge overnight, and a battery may cover part of the peak. Scheduled separately, they can all start at 4 a.m. or all restart at 10 a.m.
Three rules keep them from working against each other. Stagger the load-ups so the water heater and the heat pump’s pre-heat do not run their biggest draws in the same 15 minutes on a demand rate. Use battery power in the peak for the loads that cannot coast, not for a heat pump that could have pre-heated. And recover in steps after each peak, so the house, the water heater and the car do not all restart at once. The water heater scheduling guide gives matching load-up times for the same LG&E windows.
Culni is a home energy software company that reads a home’s electricity rate and runs one daily schedule across solar, a battery, EV charging and flexible loads, including HVAC pre-heating and pre-cooling, and then checks the result against the bill each month. Whether a particular thermostat can be scheduled that way is confirmed home by home. The all-electric homes page shows how the loads fit together.
How to tell whether it worked
A month of bills is a blunt instrument, because weather moves the total more than any schedule. Compare like with like instead.
- Use your smart meter’s hourly data if your utility offers a usage download. Compare the kWh in the peak window on similar-temperature weekdays before and after the change.
- Check the thermostat’s runtime report, if it has one; ENERGY STAR certified smart thermostats on heat pumps must report resistance heat use. Look for cooling hours that moved ahead of the peak, and for any auxiliary heat during a winter pre-heat.
- Watch the hours after the peak. If usage jumps right after the window ends, the recovery is too fast; split it into steps.
- Note the overrides. If someone in the house keeps changing the setpoint during the peak, the float is too large. The PNNL study saw overrides rise with longer events; a smaller, steadier setting beats a deep one that gets cancelled.
- Recalculate with the formula. Multiply the kWh that left the peak by the price gap, subtract the extra kWh at the off-peak price, and compare that with what the bill shows.
What to do next
- Read your rate’s windows and prices, including whether the clock is standard or daylight time and whether winter has its own peaks.
- Work out the ratio: peak price minus off-peak price, divided by off-peak price. Below about 0.2, pre-conditioning is unlikely to pay; above about 0.8, it can in a reasonably tight house.
- Run the coast test on a hot weekday, and again in winter if your rate has winter peaks.
- Set a simple weekday program: 2 to 3°F cooler for two to three hours before the peak, 3°F warmer during it, and recovery in two steps after it.
- In winter, step the pre-heat 1°F at a time, keep overnight setbacks small, and ask a technician to confirm the auxiliary heat lockout.
- Check the bill after a month against the same month’s weather, and adjust the depth and timing.
To estimate the monthly value of shifting kWh out of the peak, the time-of-use savings calculator uses your monthly use, peak share and prices; its default example, 900 kWh a month with 25% in the peak and 40% of that shifted at fictional prices of 31 and 8 cents, gives $20.70 a month. Remember to subtract the energy penalty from whatever it shows. If you would rather have the thermostat, water heater and EV windows worked out together, you can send your rate plan, a recent bill and your device list for a written home energy plan.
