Here’s the difference between building the right system and buying a pile of equipment that doesn’t work: numbers. Real numbers from your real home. Not industry averages, not a salesman’s estimate — your actual usage, your actual site, your actual sun.

This is where you get those numbers. It’s not complicated. It just takes a little time and a couple of cheap tools. And once you see how your home actually uses electricity, you’ll wonder why you ever just paid the bill without looking.


Your Bill Is the Starting Point

Pull 12 months of electric bills. Every utility has this data online — log in and get it. You’re paying for it. You’re entitled to see it.

Monthly kWh consumption. Write down all 12 months. You’ll see a pattern — higher in summer and winter, lower in spring and fall. That pattern tells you when the monopoly has the most leverage over you. It also tells you where solar can hit hardest.

Your rate structure. Are you on a flat rate, time-of-use (TOU), or tiered pricing? This matters for calculating exactly what your solar production is worth. On a TOU plan, a kWh you avoid buying during peak hours — typically 4-9 PM — is worth more than a 2 AM kWh. A battery-paired system can play that spread in your favor. See the rate calculator to see what the monopoly actually charges you, hour by hour.

Average daily usage. Monthly kWh divided by 30. That’s your baseline — the number you’re aiming to stop buying from the utility.


Measure Your Actual Loads

Your bill tells you what you spent. Now find out where the money went.

Kill-A-Watt meter (~$35). This device plugs between any appliance and the wall outlet. It measures watts in real time and tracks kWh over time. Plug your fridge in for 24 hours and you’ll know exactly what it draws — not what the manufacturer claims, not what a website guesses, but what your fridge actually pulls in your house.

Emporia Vue energy monitor (~$150). Clips onto your breaker panel and logs usage per circuit, 24/7. More money up front, but it gives you the complete picture — every circuit at once, with historical data. If you want to see where every dollar is going, this is the tool.

Nameplate vs. reality. Every appliance has a nameplate rating — the maximum it can draw. Most devices don’t run anywhere near that most of the time. A refrigerator rated at 800W might draw 65W in normal operation. That 800W is the compressor startup spike combined with a defrost heater cycle. The nameplate is the worst-case ceiling, not the day-to-day draw.

This distinction matters because you need both numbers. The typical draw tells you how much energy the device uses over time — that sizes your panels and batteries. The spike tells you the most it can demand at any instant — that sizes your inverter.

Devices aren’t constant. Your fridge hums along at 65W, then the compressor kicks in and it spikes to 200W, then twice a day the defrost cycle fires and it jumps to 800W for 20 minutes. A furnace fan runs at 250W but surges to 500+ on startup. A window AC unit draws 400W steady but needs 1,200W for the first half-second when the compressor fires. These spikes are brief — but your inverter has to handle them.


Build Your Load Profile

List every device or circuit you want to run on solar. For each one:

Continuous watts — what it draws while running normally. This is your Kill-A-Watt number.

Daily watt-hours — continuous watts times hours per day it runs. A 65W fridge running 24 hours uses 1,560 Wh (1.56 kWh) per day. A 250W furnace fan running 8 hours uses 2,000 Wh (2 kWh).

Peak spike watts — maximum instantaneous demand. Compressor starts, defrost cycles, motor startups.

Think about realistic combinations. Your peak demand isn’t every device spiking at the same instant — that almost never happens. It’s the realistic worst case of what actually runs simultaneously. A teakettle and a microwave running at the same time at breakfast — about 2,800W — is real. A washing machine and a teakettle at the same moment? Probably not. Think through your actual daily patterns.

Add it all up. Three numbers come out:


Know Your Sun

You know what you need from the system. Now find out what the sun can deliver.

Peak Sun Hours (PSH). The most important solar planning number — and the most widely misunderstood. PSH is not hours of daylight. It’s total solar energy delivered to your location, expressed as equivalent hours of perfect full-intensity sun. Five PSH could mean five hours of blazing clear sky, or it could mean ten hours of partly cloudy sky that adds up to the same total energy. What matters is the total.

PVWatts (pvwatts.nrel.gov). Free tool from the National Renewable Energy Laboratory. Enter your address and it returns monthly solar resource estimates specific to your location — your latitude, your local weather, your typical cloud cover. Don’t guess your solar potential when you can look it up for free. Run your numbers with real data.

General PSH guidance (use PVWatts for your specific location):

The reference table below shows month-by-month Portland-area numbers as an example.

Walk your property. Go outside on a sunny day and look at where you’d put panels. Track shadows — from trees, neighboring houses, chimneys, fences — at different times of day. Morning shade from an east-side tree might not matter if your panels face south. A tree directly to the south is a real problem.

Shade isn’t necessarily a dealbreaker, but it affects your design. You can isolate shaded panels on a separate string with its own MPPT input so one bad panel doesn’t drag down your whole array. Details in String Design.

Orientation. South-facing is best. East or west installations produce roughly 80% of equivalent south-facing output — that means a couple of extra panels to match the same production. Small price for a workable roof.


Can I Shrink This?

You have your load profile. Before you start pricing equipment, run one more pass at the load side.

This is “Insulate Before You Generate” made concrete. Look at your list and ask what can come down:

Iterate. Adjust the load profile, recalculate daily kWh, see how it moves the system size. This loop is free. The equipment is not.


Napkin Math

Three inputs: daily kWh target, peak load, local PSH. Three formulas. First-pass system estimate.

Panel sizing:

Daily kWh target ÷ Peak Sun Hours ÷ 0.70 = total panel watts

The 0.70 is an efficiency factor covering real-world losses — heat, wiring resistance, inverter conversion, dust, non-ideal angles. It’s conservative by design. Build in the margin and thank yourself later.

Example: 10 kWh/day ÷ 5 PSH ÷ 0.70 = 2,857W. Call it 3 kW, or seven to eight 400W panels.

This is sized for summer production. Those same panels at 1.5 PSH in December produce a fraction of that. That’s expected — design for the productive months and use storage or grid backup to bridge the winter gap.

Battery sizing:

Overnight load (watts) × hours of coverage ÷ 0.85 = minimum battery Wh

The 0.85 covers depth-of-discharge limits and round-trip efficiency losses.

Example: 500W overnight load × 10 hours = 5,000 Wh. Divided by 0.85 = approximately 5,900 Wh. Round to 6 kWh to get through a summer night. Double it for a full day of backup with no solar input.

Inverter sizing:

Peak simultaneous load × 1.25 = minimum inverter continuous rating

The 1.25 multiplier keeps your inverter out of redline territory during normal operation.

Example: 2,400W peak realistic load × 1.25 = 3,000W minimum continuous rating. Verify that the surge rating — typically 2-3x continuous — handles your largest motor-start loads.

These are ballpark numbers: “roughly a 3kW array, 6 kWh of battery, 3,000W inverter.” That’s enough to start researching real equipment. Section 3 refines these into actual specifications.

Ready to go further? Head to Panels and String Design. Or model your full system in the Energy Lab.


Reference: Common Device Wattages

Typical ranges from manufacturer data and field measurement. Measure your own devices with a Kill-A-Watt for your actual numbers.

DeviceTypical DrawSpike / PeakTypical Daily Usage
Refrigerator60-80W200-800W (compressor + defrost)1.5-2.0 kWh
Standing freezer80-100W200-500W (compressor)1.5-2.5 kWh
Furnace fan250-400W500-800W (startup)2-4 kWh (depends on run time)
Window AC (inverter)400-600W1,000-1,500W (compressor start)3-6 kWh (depends on heat/runtime)
Internet router + modem15-30Wminimal0.4-0.7 kWh
LED lights (whole house)100-300Wminimal1-2 kWh
Microwave1,000-1,200Wsame (resistive load, no spike)0.1-0.3 kWh (short use)
Electric kettle1,200-1,500Wsame (resistive)0.1-0.2 kWh (short use)
Clothes dryer4,000-5,500W (240V)same2.5-4.0 kWh per load
Electric oven2,000-6,000W (240V)samevaries widely
Washing machine300-500W500-800W (motor start)0.3-0.5 kWh per load
Laptop charging30-65Wminimal0.2-0.5 kWh
Desktop gaming PC150-400Wsame3.6-9.6 kWh (if always on)
Phone charging5-20Wminimal0.05-0.1 kWh
TV (LED, 50-65”)50-100Wminimal0.3-0.8 kWh
CPAP machine30-60Wminimal0.2-0.5 kWh
Sump pump300-800W1,500-2,500W (startup)varies (intermittent)
Well pump750-1,500W (240V)2,000-4,500W (startup)varies
EV charger (Level 1)1,200-1,400Wminimal8-12 kWh (overnight)

Note: 240V devices (dryer, oven, well pump, EV charger) require a 240V inverter. A 120V system cannot run these loads — plan accordingly or identify alternatives before finalizing your profile.


Reference: Peak Sun Hours

Monthly averages for the Portland, OR area. Use PVWatts for your specific location.

MonthPSH (avg)Notes
January1.2Shortest days, lowest sun angle, heavy clouds
February1.8Slightly better, still tough
March2.8Spring begins, noticeable improvement
April3.8Good production starts
May4.8Strong month
June5.5Peak production — longest days, highest sun
July5.8Best month — long days, clearest skies
August5.3Still excellent, days getting shorter
September4.0Solid shoulder month
October2.5Falling off
November1.5Winter begins
December1.0The bottom — plan accordingly

Use PVWatts (pvwatts.nrel.gov) for numbers specific to your address.


Previous: Define Your Goals | Next: Panels


DATA SOURCED FROM: National Renewable Energy Laboratory (NREL) — PVWatts solar resource data. Portland-area PSH values are approximate regional averages based on NREL data; individual results vary by location, orientation, and shading. Device wattage figures are representative ranges from manufacturer specifications and field measurements. Individual results vary — measure your own devices.