Every off-grid power system is sized from one number: how much energy the property uses in a day. Get it wrong and the system is either too small to carry you through winter or far more expensive than it needs to be. A careful load audit measures that number instead of guessing it, finds the loads that waste the most, and shows where a small change in equipment saves more money than extra panels or batteries ever could. This guide covers the units involved, how to measure every kind of load, duty cycles, standby power, power factor, peak demand, seasonal audits, efficiency upgrades with real numbers, and how to turn the audit into a daily energy budget you can live with. Our solar power sizing guide shows how to turn the final audit into an array, controller and inverter; this guide is about getting the audit itself right.
Power, energy and current
Three units appear in every audit, and mixing them up is the most common source of error.
- Power (watts, W): the rate at which a device uses energy at a given moment. A 9 W LED lamp uses 9 W whenever it is on.
- Energy (watt-hours, Wh, or kilowatt-hours, kWh): power multiplied by time. That 9 W lamp running for 4 hours uses 9 × 4 = 36 Wh. One kWh is 1,000 Wh.
- Current (amps, A): the flow of electricity. Power in watts equals volts × amps for a simple resistive load. A 12 V device drawing 3 A uses 36 W.
Battery capacity is often given in amp-hours (Ah). Multiply by the battery voltage to get energy: a 100 Ah battery at 12.8 V nominal holds 100 × 12.8 = 1,280 Wh. Always convert loads and batteries to watt-hours before comparing them, because an amp-hour at 12 V holds half the energy of an amp-hour at 24 V.
Power and energy size different parts of the system. Daily energy in watt-hours sets the battery bank and the solar array. Peak power in watts, including starting surges, sets the inverter. An audit needs both.
Step 1: list every load
Walk through the building room by room and outside, and write down everything that uses electricity. Include the loads that are easy to forget:
- water pumps, pressure pumps and any UV disinfection lamp;
- furnace and boiler blowers, circulator pumps and bathroom or range hood fans;
- heat trace cable on water lines;
- internet terminals, routers, cordless phone bases and security cameras;
- smoke and CO alarms wired to the electrical system;
- tool battery chargers, phone and tablet chargers left plugged in;
- the power system's own loads: inverter idle draw, charge controller, battery monitor and any remote display.
For each load, record its power, how many there are, and how many hours a day it runs in each season. A spreadsheet with one row per load and one column per month makes the seasonal audit later much easier.
Step 2: measure, do not estimate
Nameplate ratings
The nameplate on an appliance gives its maximum input, usually in watts or as volts and amps. It is a maximum, not an average. Multiplying a nameplate's volts by its amps gives the maximum apparent power, which for many motor and electronic loads is well above what the device draws in normal use. Use nameplates only for loads you cannot measure, and treat the result as a conservative estimate.
Plug-in energy meters
A plug-in energy meter sits between the outlet and the appliance and records both instantaneous watts and accumulated watt-hours. It is the most useful tool in an audit.
- For loads that cycle, such as refrigerators and freezers, record accumulated watt-hours over at least 24 hours, and longer if door openings vary through the week.
- For loads used irregularly, such as a coffee maker or a printer, measure one typical use and multiply by uses per day.
- For devices that are always plugged in, read the watts with the device off. That is its standby draw.
Clamp meters for hard-wired loads
Well pumps, furnace blowers and other hard-wired loads cannot use a plug-in meter. A clamp meter placed around one conductor of the circuit reads current without disconnecting anything. Multiply by the circuit voltage to get apparent power in volt-amperes (VA). For motors, true power in watts is lower by the power factor, which a basic clamp meter cannot show, so the VA figure is an upper bound. Time how long the load runs per cycle and how many cycles occur per day.
Battery monitors and system logs
A battery monitor with a shunt on the battery negative cable measures every amp in and out of the bank, including DC loads and everything running through the inverter. It is the most accurate whole-system meter, and it already includes inverter losses. Many inverter-chargers and charge controllers also log daily energy.
Use the monitor for a subtraction test: switch off every load you can, note the remaining draw, then switch each load on in turn and note the increase. The remaining draw with everything off is your base load, and it is often larger than owners expect.
Duty cycles
Many loads switch on and off by themselves. The fraction of time a load runs is its duty cycle.
Daily energy (Wh) = running power (W) × duty cycle × 24 hours
Worked example: finding a refrigerator's duty cycle
A plug-in meter shows a refrigerator's compressor drawing 90 W when it runs. Over 24 hours the meter records 1,200 Wh.
Average power is 1,200 ÷ 24 = 50 W. The duty cycle is 50 ÷ 90 = 0.56, so the compressor runs about 56% of the time.
If you had estimated the refrigerator at its 90 W running power for 24 hours, you would have entered 2,160 Wh, an overestimate of 960 Wh a day. If you had guessed a 30% duty cycle, you would have entered 648 Wh, an underestimate of 552 Wh. Measuring removes both errors.
A refrigerator's duty cycle rises with room temperature, door openings and warm food going in, and falls when the room is cool. Measure it in the season that matters for sizing, and in the location where it will actually sit. Natural Resources Canada advises keeping refrigerators away from ovens, heating vents and direct sun, cleaning the condenser coils regularly and not overfilling the compartments, all of which shorten run time.
Standby and always-on loads
Standby power is drawn by devices that are switched off but still plugged in, and by devices that never switch off, such as routers. Natural Resources Canada says standby power can account for 5 to 10% of a household's electricity bill. Hydro-Québec, citing Natural Resources Canada, notes that the average home has 20 to 40 consumer electronic devices drawing phantom power, and that a device in standby may draw as little as 0.5 W or almost 30 W.
On the grid this is a modest cost. Off-grid it matters far more, because an always-on load runs 24 hours a day, including through the long nights of December when the battery has to carry it.
Worked example: the cost of small loads
A cabin's subtraction test shows these always-on loads:
- television and sound bar in standby: 6 W
- game console in standby: 5 W
- microwave clock and display: 3 W
- three chargers left plugged in: 4 W
Total: 6 + 5 + 3 + 4 = 18 W. Over a day that is 18 × 24 = 432 Wh, which is what a 90 W refrigerator compressor uses running nonstop for 4.8 hours.
Putting the television, console and microwave on a switched power strip, and unplugging chargers when not in use, removes almost all of it.
Regulation has helped. Televisions and compact audio products sold in Canada are regulated to draw no more than 0.5 W when off, so a modern set uses far less in standby than an older one. Network-connected devices are the exception: Natural Resources Canada notes that some use nearly as much power in standby as when fully on. Hydro One adds that phone and tablet chargers can lose up to half the electricity they draw as heat.
Power factor and inverter ratings
For a heater or an incandescent lamp, volts × amps equals watts. For motors and many electronic power supplies, current and voltage are out of step, and volts × amps is larger than the real power. The ratio of real power (W) to apparent power (VA) is the power factor, a number between 0 and 1.
- Batteries and solar arrays deliver real energy, so the energy audit is always done in watts and watt-hours.
- The inverter must also supply the apparent power. If an inverter is rated in VA, or has both a VA and a W rating, check that the loads running together fit within both.
- Power factor also affects current. A load with a low power factor draws more amps than its wattage suggests, which matters for wire and breaker sizing on the AC side.
A plug-in meter that displays watts, VA and power factor shows these figures directly for each appliance. Record all three for motor loads.
Peak demand and load management
Daily energy tells you nothing about what happens at 7 a.m. when the pump, toaster and lights all come on. The audit needs a second list: which loads might run at the same moment.
Worked example: the morning peak
On a winter morning these loads could all run together:
- well pump: 800 W running
- toaster: 900 W
- refrigerator compressor: 90 W
- lights: 6 × 9 W = 54 W
- laptop: 60 W
- internet terminal: 30 W
Running total: 800 + 900 + 90 + 54 + 60 + 30 = 1,934 W. If the pump's motor starts while everything else is already running, the inverter must also supply the pump's starting surge on top of the other 1,134 W.
Moving toast to after the pressure tank has refilled cuts the peak by 900 W. Simple habits like this can let a smaller, cheaper inverter carry the household.
Write a short load management plan and post it near the inverter:
- which large loads should never run together;
- which loads to switch off first when the battery is low, starting with the least essential;
- which loads to save for sunny afternoons or generator runs, such as laundry, power tools and vacuuming.
Seasonal audits
An off-grid property's use changes with the seasons and with occupancy. A single annual average hides the month that decides the system size.
| Load | Why it changes | Highest in |
|---|---|---|
| Lighting | Longer nights | December |
| Furnace or boiler blower, circulators | Heating season | January |
| Heat trace | Only below freezing | January and February |
| Refrigeration | Warmer rooms raise duty cycle | July and August |
| Water pumping | Garden watering, more guests | Summer |
| Fans and ventilation | Cooling and moisture control | Summer |
Fill in the spreadsheet for every month you occupy the property. Then compare each month's daily load with the solar energy available in that month. The month with the worst ratio of load to sun is the design month; for most northern properties it is December or January, but a summer-only cabin with heavy refrigeration and pumping may have its design month in July.
Occupancy matters as much as season. Record separate figures for weekdays and weekends, and for times with guests, if they differ significantly.
Efficiency upgrades
Every watt-hour removed from the daily load is a watt-hour you do not need to generate, store and convert. Work through the list from the biggest saving per dollar.
Lighting
Natural Resources Canada states that ENERGY STAR certified bulbs use up to 90% less energy than traditional incandescent bulbs and that certified LEDs last at least 15 times longer. Replacing any remaining incandescent or halogen lamp is the cheapest saving in most audits. Choose 12 V or 24 V DC LED fixtures for lights on a dedicated DC circuit to avoid inverter losses.
Refrigeration
Natural Resources Canada gives these figures:
- ENERGY STAR refrigerators use 9% less energy, on average, than a standard model.
- An old refrigerator uses about 35% more energy, on average, than an ENERGY STAR model.
- When you replace a refrigerator, recycle the old one rather than moving it to a garage or cottage.
- Keep the refrigerator at 1.7 to 3.3 °C (35 to 38 °F) and the freezer at −18 °C (0 °F). Colder settings waste energy.
- Replace door seals if a sheet of paper closed in the door slides out easily.
Compare models using the annual kilowatt-hours on the EnerGuide label; a lower number is better. Divide by 365 for a daily figure: a model labelled 350 kWh per year averages 350 ÷ 365 = 0.96 kWh, or 960 Wh, per day by the label figure. Measure it after installation, since your room temperature and use will differ.
Internet and electronics
Satellite internet is now a major continuous load at many cabins. Starlink's specification sheets list average power consumption of 35 to 50 W for its standard (V5) terminal and 25 to 40 W for the Mini, which accepts 12 to 48 V DC input. Running a terminal directly from the battery through a suitable DC supply avoids inverter losses, and switching it off overnight with a timer removes a third of its daily energy.
Pumps and water
Pump energy is proportional to the water pumped. Every litre saved by low-flow fixtures, a low-flush toilet or fixing a leak also saves pumping energy. Measure the pump's watt-hours per 100 L once, then use it to price water savings in watt-hours.
Fuel switching
Heat-producing loads, such as space heaters, water heaters, dryers and electric ranges, are covered in our solar sizing guide. If any appear in your audit, they are almost always better moved to propane, wood or another fuel than supplied with solar.
What a saved watt-hour is worth
The design factors in our solar sizing guide (a system efficiency of 0.85, a panel derate of 0.85 and, in the central Ontario example, 1.8 December peak sun hours) let you price a saving in panels and batteries.
Worked example: replacing six incandescent lamps
Six 60 W incandescent lamps run 4 hours a night: 6 × 60 × 4 = 1,440 Wh per day. Six 9 W LEDs giving similar light use 6 × 9 × 4 = 216 Wh. The saving is 1,440 − 216 = 1,224 Wh per day.
Array avoided: 1,224 ÷ 0.85 ÷ (1.8 × 0.85) = 1,224 ÷ 0.85 ÷ 1.53 = 941 W of panels, before any recovery margin.
Battery avoided: with 92% inverter efficiency, two days of autonomy and 80% usable depth of discharge, 1,224 ÷ 0.92 × 2 ÷ 0.80 = 3,326 Wh of battery.
Six LED lamps cost a small fraction of nearly a kilowatt of panels and more than 3 kWh of batteries. That comparison holds for almost every efficiency upgrade in an audit.
The daily energy budget
Once the system is built, the audit becomes a budget. Divide the system's design daily output among the household's needs, and watch the battery monitor to check that real use matches.
- Read the battery monitor's daily energy figure at the same time each day for the first month, and write it down with the weather.
- Compare each day with the audit. If use is consistently higher, repeat the subtraction test to find the load that is larger than expected.
- Keep a running total through cloudy spells. When the battery's state of charge falls below your chosen threshold, apply the load management plan or run the generator.
- Repeat the full audit after any major change: a new appliance, a new pump, a change in occupancy or a new season of use.
Worked example: a full audit before and after
A weekend cabin's first audit, measured with a plug-in meter, clamp meter and battery monitor:
| Load | Before (Wh per day) | Change | After (Wh per day) |
|---|---|---|---|
| Lighting | 1,440 (6 × 60 W × 4 h) | LED lamps | 216 |
| Old refrigerator | 1,600 | ENERGY STAR model (old uses about 35% more) | 1,600 ÷ 1.35 = 1,185 |
| Internet terminal | 1,080 (45 W × 24 h) | Smaller terminal at 30 W, 16 h a day | 480 |
| Standby loads | 432 (18 W × 24 h) | Switched power strip; one 2 W device remains | 48 |
| Well pump | 240 (800 W × 0.3 h) | None | 240 |
| Laptop | 180 (60 W × 3 h) | None | 180 |
| Total | 4,972 | 2,349 |
The changes cut daily use by 4,972 − 2,349 = 2,623 Wh, or 53%.
Using the same December design factors, the break-even array falls from 4,972 ÷ 0.85 ÷ 1.53 = 3,823 W to 2,349 ÷ 0.85 ÷ 1.53 = 1,806 W, less than half the size.
Key points
- Work in watt-hours for energy and watts for peak demand; convert battery amp-hours to watt-hours by multiplying by voltage.
- List every load, including pumps, blowers, heat trace, internet, alarms, chargers and the power system's own draw.
- Measure with a plug-in meter over at least 24 hours for cycling loads, a clamp meter for hard-wired loads, and a battery monitor for the whole system.
- Standby power can be 5 to 10% of a household's electricity bill; off-grid, an 18 W always-on load costs 432 Wh a day.
- Check inverter ratings against both watts and volt-amperes, and plan which large loads never run together.
- Audit every month you occupy the property and size for the month with the worst ratio of load to sun.
- Upgrade lighting, refrigeration and always-on electronics before buying power equipment; each saved watt-hour avoids panels and batteries.
This guide is general information. Codes, permits and safety requirements vary by province, state and municipality. Check local rules and use a qualified professional for electrical, gas, chimney and septic work where required.
The Comprehensive Off-Grid Manual
Every off-grid system in one manual, with worked examples, checklists and northern-climate detail.
More guides
Off-Grid Solar Power: How to Size a System
Size an off-grid solar system for Canadian winters: load audit, peak sun hours, array, MPPT, inverter, wiring, fusing, with a full cabin example.
Off-Grid Batteries: LiFePO4 vs Lead-Acid
LiFePO4 vs lead-acid for off-grid power: usable capacity, cycle life, cold-weather charging, bank sizing, wiring, ventilation and lifetime cost.
Generators: Sizing, Fuel and Safe Operation
Choose and size an off-grid generator, compare gasoline, propane and diesel, follow fuel storage limits, plan fuel use, prevent CO poisoning and connect safely.